old/paleo_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 134,
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import random\n",
    "from sklearn.metrics import explained_variance_score, \\\n",
    "    mean_absolute_error, \\\n",
    "    median_absolute_error\n",
    "from sklearn.model_selection import train_test_split"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "source": [
    "np.random.seed(4747)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Let's run through a basic analysis of one proxy Organism"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "#### Arctic Bay, Arctic: Coralline Algae (1967-2008)"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "source": [
    "raw_paleo_data = pd.read_csv(\"d18O_raw/ab_d18O_raw.csv\")\n",
    "raw_paleo_data.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "   year      d18O\n",
       "0  2008 -0.048160\n",
       "1  2007 -0.045444\n",
       "2  2006 -0.070624\n",
       "3  2005 -0.095804\n",
       "4  2004  0.027000"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>d18O</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2008</td>\n",
       "      <td>-0.048160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2007</td>\n",
       "      <td>-0.045444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2006</td>\n",
       "      <td>-0.070624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2005</td>\n",
       "      <td>-0.095804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2004</td>\n",
       "      <td>0.027000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 136
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "source": [
    "temp_data = pd.read_csv(\"temp_raw/ab_SODA_1x1_raw.csv\")\n",
    "temp_data.head(10)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "   Unnamed: 0                  time    lat     lon  depth      temp       salt\n",
       "0           1  1967-01-15T00:00:00Z  72.75  273.75   5.01 -1.280347  31.575184\n",
       "1           2  1967-01-15T00:00:00Z  72.75  274.25   5.01 -1.280931  31.576525\n",
       "2           3  1967-01-15T00:00:00Z  73.25  273.75   5.01 -1.298791  31.579832\n",
       "3           4  1967-01-15T00:00:00Z  73.25  274.25   5.01 -1.324163  31.503399\n",
       "4           5  1967-01-15T00:00:00Z  73.75  273.75   5.01 -1.425924  30.357185\n",
       "5           6  1967-01-15T00:00:00Z  73.75  274.25   5.01 -1.409098  30.414881\n",
       "6           7  1967-02-15T00:00:00Z  72.75  273.75   5.01 -1.279240  31.829067\n",
       "7           8  1967-02-15T00:00:00Z  72.75  274.25   5.01 -1.281839  31.826092\n",
       "8           9  1967-02-15T00:00:00Z  73.25  273.75   5.01 -1.320389  31.776814\n",
       "9          10  1967-02-15T00:00:00Z  73.25  274.25   5.01 -1.360575  31.597593"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>time</th>\n",
       "      <th>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>depth</th>\n",
       "      <th>temp</th>\n",
       "      <th>salt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.280347</td>\n",
       "      <td>31.575184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.280931</td>\n",
       "      <td>31.576525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.298791</td>\n",
       "      <td>31.579832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.324163</td>\n",
       "      <td>31.503399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.75</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.425924</td>\n",
       "      <td>30.357185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.75</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.409098</td>\n",
       "      <td>30.414881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>1967-02-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.279240</td>\n",
       "      <td>31.829067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>1967-02-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.281839</td>\n",
       "      <td>31.826092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>1967-02-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.320389</td>\n",
       "      <td>31.776814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>1967-02-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.360575</td>\n",
       "      <td>31.597593</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 137
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Let's make sure this temperature is averaged across the grid such that each month only has one temperature and one salinity value across the multiple lat/long pairs. "
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "source": [
    "monthly_temp_data = temp_data.groupby('time')\n",
    "\n",
    "print(monthly_temp_data['temp'].agg(np.mean))"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "time\n",
      "1967-01-15T00:00:00Z   -1.336542\n",
      "1967-02-15T00:00:00Z   -1.343062\n",
      "1967-03-15T00:00:00Z   -1.377000\n",
      "1967-04-15T00:00:00Z   -1.455148\n",
      "1967-05-15T00:00:00Z   -1.514044\n",
      "                          ...   \n",
      "2008-08-15T00:00:00Z   -0.197338\n",
      "2008-09-15T00:00:00Z   -0.944179\n",
      "2008-10-15T00:00:00Z   -1.022546\n",
      "2008-11-15T00:00:00Z   -1.116120\n",
      "2008-12-15T00:00:00Z   -1.164364\n",
      "Name: temp, Length: 504, dtype: float64\n"
     ]
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Let's average these temperatures across years, across the grids. "
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "source": [
    "temp_data['year'] = temp_data['time'].apply(lambda x: x[0:4])\n",
    "temp_data"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "      Unnamed: 0                  time    lat     lon  depth      temp  \\\n",
       "0              1  1967-01-15T00:00:00Z  72.75  273.75   5.01 -1.280347   \n",
       "1              2  1967-01-15T00:00:00Z  72.75  274.25   5.01 -1.280931   \n",
       "2              3  1967-01-15T00:00:00Z  73.25  273.75   5.01 -1.298791   \n",
       "3              4  1967-01-15T00:00:00Z  73.25  274.25   5.01 -1.324163   \n",
       "4              5  1967-01-15T00:00:00Z  73.75  273.75   5.01 -1.425924   \n",
       "...          ...                   ...    ...     ...    ...       ...   \n",
       "3019        3020  2008-12-15T00:00:00Z  72.75  274.25   5.01 -1.109940   \n",
       "3020        3021  2008-12-15T00:00:00Z  73.25  273.75   5.01 -1.116151   \n",
       "3021        3022  2008-12-15T00:00:00Z  73.25  274.25   5.01 -1.128340   \n",
       "3022        3023  2008-12-15T00:00:00Z  73.75  273.75   5.01 -1.263921   \n",
       "3023        3024  2008-12-15T00:00:00Z  73.75  274.25   5.01 -1.255458   \n",
       "\n",
       "           salt  year  \n",
       "0     31.575184  1967  \n",
       "1     31.576525  1967  \n",
       "2     31.579832  1967  \n",
       "3     31.503399  1967  \n",
       "4     30.357185  1967  \n",
       "...         ...   ...  \n",
       "3019  31.780386  2008  \n",
       "3020  31.646585  2008  \n",
       "3021  31.521568  2008  \n",
       "3022  30.873104  2008  \n",
       "3023  30.926878  2008  \n",
       "\n",
       "[3024 rows x 8 columns]"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>time</th>\n",
       "      <th>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>depth</th>\n",
       "      <th>temp</th>\n",
       "      <th>salt</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.280347</td>\n",
       "      <td>31.575184</td>\n",
       "      <td>1967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.280931</td>\n",
       "      <td>31.576525</td>\n",
       "      <td>1967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.298791</td>\n",
       "      <td>31.579832</td>\n",
       "      <td>1967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.324163</td>\n",
       "      <td>31.503399</td>\n",
       "      <td>1967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>1967-01-15T00:00:00Z</td>\n",
       "      <td>73.75</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.425924</td>\n",
       "      <td>30.357185</td>\n",
       "      <td>1967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3019</th>\n",
       "      <td>3020</td>\n",
       "      <td>2008-12-15T00:00:00Z</td>\n",
       "      <td>72.75</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.109940</td>\n",
       "      <td>31.780386</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3020</th>\n",
       "      <td>3021</td>\n",
       "      <td>2008-12-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.116151</td>\n",
       "      <td>31.646585</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3021</th>\n",
       "      <td>3022</td>\n",
       "      <td>2008-12-15T00:00:00Z</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.128340</td>\n",
       "      <td>31.521568</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3022</th>\n",
       "      <td>3023</td>\n",
       "      <td>2008-12-15T00:00:00Z</td>\n",
       "      <td>73.75</td>\n",
       "      <td>273.75</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.263921</td>\n",
       "      <td>30.873104</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3023</th>\n",
       "      <td>3024</td>\n",
       "      <td>2008-12-15T00:00:00Z</td>\n",
       "      <td>73.75</td>\n",
       "      <td>274.25</td>\n",
       "      <td>5.01</td>\n",
       "      <td>-1.255458</td>\n",
       "      <td>30.926878</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3024 rows × 8 columns</p>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 139
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Voila, now we have the average Sea Surface Temp for that lat/long grid on an annual basis:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "source": [
    "actual_annual_temp = temp_data.groupby('year')['temp'].agg(np.mean).to_frame().reset_index()\n",
    "# annual_temp['lat'] = 73.25\n",
    "# annual_temp['long'] = 274.75\n",
    "\n",
    "actual_annual_temp.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "   year      temp\n",
       "0  1967 -1.301917\n",
       "1  1968 -0.175512\n",
       "2  1969 -0.575391\n",
       "3  1970 -1.314653\n",
       "4  1971 -1.344812"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>temp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1967</td>\n",
       "      <td>-1.301917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1968</td>\n",
       "      <td>-0.175512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1969</td>\n",
       "      <td>-0.575391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>-1.314653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1971</td>\n",
       "      <td>-1.344812</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 140
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Let's see how these two factors naturally correlate:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "source": [
    "#pre-process for the merge\n",
    "\n",
    "actual_annual_temp['year'] = actual_annual_temp['year'].apply(lambda x: int(x))"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "source": [
    "temp_and_algae = pd.merge(actual_annual_temp, raw_paleo_data, on='year')\n",
    "temp_and_algae.set_index('year').head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "          temp      d18O\n",
       "year                    \n",
       "1967 -1.301917 -0.132046\n",
       "1968 -0.175512 -0.134891\n",
       "1969 -0.575391 -0.137737\n",
       "1970 -1.314653 -0.134845\n",
       "1971 -1.344812 -0.131952"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>temp</th>\n",
       "      <th>d18O</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1967</th>\n",
       "      <td>-1.301917</td>\n",
       "      <td>-0.132046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1968</th>\n",
       "      <td>-0.175512</td>\n",
       "      <td>-0.134891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1969</th>\n",
       "      <td>-0.575391</td>\n",
       "      <td>-0.137737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1970</th>\n",
       "      <td>-1.314653</td>\n",
       "      <td>-0.134845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1971</th>\n",
       "      <td>-1.344812</td>\n",
       "      <td>-0.131952</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 142
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "source": [
    "import matplotlib.pyplot as plt\n"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "source": [
    "plt.plot(temp_and_algae['year'], temp_and_algae['temp'])\n",
    "plt.plot(temp_and_algae['year'], temp_and_algae['d18O'])\n",
    "plt.title('Arctic Bay: Temperature and D18O')\n",
    "plt.xlabel('Year')\n",
    "plt.legend(['temp', 'd18O'], loc='upper right')\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "We can see from the graph that there's some sort of relationships between the spikes in temp and d18O, but our correlation coefficient is rather underwhelming:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "source": [
    "np.corrcoef(temp_and_algae['temp'], temp_and_algae['d18O'])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([[ 1.        , -0.12974504],\n",
       "       [-0.12974504,  1.        ]])"
      ]
     },
     "metadata": {},
     "execution_count": 145
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "source": [
    "import statistics\n",
    "statistics.variance(temp_and_algae['temp']), statistics.variance(temp_and_algae['d18O'])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(0.19341078570023343, 0.015289237992896332)"
      ]
     },
     "metadata": {},
     "execution_count": 146
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "source": [
    "plt.scatter(temp_and_algae['temp'], temp_and_algae['d18O'])\n",
    "plt.title('Arctic Bay: Temperature and D18O')\n",
    "plt.xlabel('temp')\n",
    "plt.ylabel('d18O')\n",
    "\n",
    "# plt.legend(['temp', 'd18O'], loc='upper right')\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEWCAYAAAB42tAoAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAfsUlEQVR4nO3de7gcdZ3n8feHEOQg6AFBISeJiZgFwRDQswhGRxTYAI4mkx2viIy3LKOiuGM06KMTdN1kH1xvq4IZVOIFL6OZGASNSryMIpJAlKuRAEpuQAiEi0RJyHf/qN+R5tCnuk+f7q7u6s/refpJd1V11fd0Vepbv0v9ShGBmZnZSPYoOgAzM+tsThRmZpbLicLMzHI5UZiZWS4nCjMzy+VEYWZmuZwobNQkXSjpQ0XHYd1H0gmSNhYdh42OE0WJSfqZpPskPWkM6/gnSb+snBYRZ0XERxtY1x8l7ZD0UIrrMkmTGo2tju1NTtsaeoWkP1d8fnGrtt1K6e94dtFxtMKwfbRN0hWSXjNsmVdLulLSw5J+VmUdL5N0raQHJN0mad6w+UdIWiHpfkkPSvqppBe2+E/rak4UJSVpCvBiIIBX1lh2XDtiSl4REfsChwB3Af+vVRuKiDsiYt+hV5o8o2Laf7Zq242StGcbttHO/d2IGWl/HQZcDHxW0r9WzL8X+BSwePgXJY0H/gP4AvBU4DXAJyTNSPMPBX4FXA9MBSak5X8k6fgW/T3dLyL8KuEL+DDZf4hPAN8fNu9i4ALgcuDPwEnAJGAZsBXYBnwWeA7wF+BR4CFge8X3/1fF+mYDvwUeAG4FThkhpj8CJ1V8Pg34Q8XnlwNr03o2AAsr5l0GnD1sfdcBc0bxmwTw7PT+ScDHgTvIEtaFQF+adwKwEXgfcDewBZgzFC/ZieoDFetdCHwH+BbwIHAt2cluaP4E4Lvpt70deFeV734t/d1vBY4Ffg1sT9v+LLBXWv4X6e/4c9onrwH+Cfhlzt9abX+PGFOV3y1vv0xJ2zoz/Zb3AB+smN+Xtn8fcBMwH9hYzz6qmPaPZMfh04ZNfyvws2HTnpHWsU/FtNXA69L7rwKXV9nuBcAviv5/26mvwgPwq0U7FtYDbweeD+wEnlEx72LgfmAmWanyycDvgE+m93sDL0rLVjsJXUxKFOmkdj9wclrXAHD4CDH9kZQogH2ApcBXKuafAExP6zmK7AQ+J817NfCbimVnkCW0vYDPA5+v4zepPHl+ClgBHADsB1wKLKqIYxdZsh0PvC2dUC9Jyx6ZTlzPSssvTL/xP6bl30t28h2f/pZr0rr2Ap4F3AbMGvbdOWnZvrTPjgP2JDsR3wycU+3vyNlHwxNF5f7eJy+mKr9b3n6Zkrb1byn2GcBfgeek+YuB/0y/8yTgBkafKMan/XHqsOlPSBRp+iXAO4BxwPFkyX5Smncn8KYq33kp2QXRPiPF1suvwgPwqwU7FV6UTj4Hps+/B95TMf9iHn+CPj6dCPessq5qJ6GLeSxRfAH4ZJ1x/ZFUMkn/8TcD03OW/9TQuslKAPcC09Lnj1NHchi2vgCeDYjsyvrQYb/B7en9CcAOYFz6vF/67gsqlr+m4mS5ELiqYt4eZCWBFwMvAO4YFse5wJcrvpt7JQucA/zH8L+jxj4anigq93duTHX8jpX7ZUra1sSK+VcDr03vb6OihAnMY5SJIk2/Ezh92LSREsUryJLZrvR6W8W8XVQp8QKHp20PjOaY6pWX2yjK6UzgRxFxT/p8SZpWaUPF+0nAnyJiVwPbmkRW3VSvORHRT3bifyfwc0kHA0h6QWpY3CrpfuAs4ECAiPgr8G3gDZL2AF5HVo3QiINIV9WStkvaDvwwTR+yLSIeTe93pH/vqpi/A9i34vPffs+I2E1WdTUBeCYwYWg7aVsfIKsiecJ3AST9F0nfl3SnpAeA/036Hcagchv1xFQZz4j7pcKdFe8f5rHfZsKwbf9ptIGndoeDyC4Uai17OFkV4BvJSktHAu+T9PK0yD1k7WPDHQLsJqsis2GcKEpGUh9ZNc1L0onmTuA9wIyhBr0kKt5vACaP0JAaVaZV2gAcOto4I+LRiFhGVtx/UZp8CVl10KSIeCpZu4EqvrYUOB04EXg4In492u0m95Cd6I+MiP70emo81uDdiL/13kqJbCJZiWkDWUmlv+K1X0ScVvHd4b/xBWSlwGkR8RSyk7gY2Z/JEt/Q9g+usszw/V0rpkq19kueLVT8NsDkOr9XaTZZSeDqOpZ9LrAuIlZGxO6IWEfWvnVqmv8T4FVVvvdq4NcR8XAD8ZWeE0X5zCE7+R4BHJ1ezyGrJ37jCN+5muw/9GJJT5a0t6SZad5dwERJe43w3S8Cb5J0oqQ9JA2kq7pcyswG9ierg4esiufeiPiLpGOB11d+JyWG3cD/pfHSxNAV/78Bn5T09BTPgKRZja4TeL6kuSnZnkNWT38V2W/7gKT3S+qTNE7ScyX915x17UfWcPxQ+i3/edj8u8jaFYb8DjhS0tGS9iarzsoz2phy90sN3wbOlbS/pInA2fV+UdIBkk4HPgf8n4jYlqaPS3/nnsAe6Xgdn762FpiWusgq9XL6e7LfCOA84IWSPpbWv5+ks8n+b7x/FH9XT3GiKJ8zyeqa74iIO4deZD1nTq9WakhVLK8gq7+/g6zaZKjv+irgRuBOSfdU+e7VwJvIGsLvB35OVrUxkkslPUR2IvwYcGZE3JjmvR34iKQHyRpav13l+18ha1j92tAEZTcAXpizzWreT9bgf1Wq3vkJWXfMRn2P7De7DzgDmBsROyt+26PJGrjvAS4i67o5kveSnYwfJEto3xo2fyGwNFUbvToi/gB8JP0NtwC/JEcDMdWzX0ZyHll10+3Aj6gvwf8uHSPrydoh3hMRH66YfwZZifACsnagHWS/ExFxK/Bm4DNkx9jPyXp3fTHNv4WsBDuDrM1sC/DfyRryfzWKv6unKDXkmHUFSW8E5kXEi2ou3CaSFpI1wL6h6FjMWsElCusakvYhu7pdUnQsZr3EicK6Qmo/2EpWP39JweGY9RRXPZmZWS6XKMzMLFfLByArwoEHHhhTpkwpOgwzs65xzTXX3BMRB1WbV8pEMWXKFNasWVN0GGZmXUPSiHfNu+rJzMxyOVGYmVkuJwozM8vlRGFmZrmcKMzMLFcpez11o+VrN3H+ynVs3r6DCf19zJ91GHOOGSg6LDMzJ4pOsHztJs5ddj07dmbPydm0fQfnLrsewMnCzArnqqcOcP7KdX9LEkN27HyU81euKygiM7PHOFF0gM3bd4xquplZOzlRdIAJ/X2jmm5m1k5OFB1g/qzD6Bs/7nHT+saPY/6ssTxwrZyWr93EzMWrmLrgMmYuXsXytZuKDsms9NyY3QGGGqzd6ymfG/3NiuFE0SHmHDPgk10NeY3+/u3MWsdVT9Y13OhvVgwnCusabvQ3K4YThXUNN/qbFcNtFNY13OhvVgwnCusqbvQ3az9XPZmZWa5CE4WkUyStk7Re0oIq80+XdF16XSlpRhFxtptvKjOzTlJY1ZOkccDngJOBjcBqSSsi4qaKxW4HXhIR90k6FVgCvKD90baPbyozs05TZIniWGB9RNwWEY8A3wRmVy4QEVdGxH3p41XAxDbH2HYeSdbMOk2RiWIA2FDxeWOaNpK3AD9oaUQdwDeVmVmnKTJRqMq0qLqg9FKyRPH+EVcmzZO0RtKarVu3NinE9vNNZWbWaYpMFBuBSRWfJwKbhy8k6SjgImB2RGwbaWURsSQiBiNi8KCDDmp6sO3im8rMrNMUeR/FamCapKnAJuC1wOsrF5A0GVgGnBERf2h/iO3nm8rMrNMUligiYpekdwIrgXHAlyLiRklnpfkXAh8GngZ8XhLArogYLCrmdvFNZWbWSRRRtVmgqw0ODsaaNWuKDsPMrGtIumakC3HfmW1mZrmcKMzMLJcHBUyWr93kBmQzsyqcKBjbsBlOMGZWdk4UNP4s5m4dl8nJzcxGw20UND5sRjeOyzSU3DZt30HwWHLzCLVmNhInChofNqMbx2XqxuRmZsVyoqDxYTO6cVymbkxuZlYsJwqy9oRFc6cz0N+HgIH+PhbNnV6z3r4bx2XqxuRmZsVyY3bSyLAZ3Tgu0/xZhz2uAR46P7mZWbGcKMao28Zl6sbkZmbFcqLoQd2W3MysWE4UPcr3UphZvZwoelC33ihoZsVwr6ce5HspzGw0nCh6kO+lMLPRcKLoQb6XwsxGw4miB3XjjYJFW752EzMXr2LqgsuYuXiVx8aynuLG7B7keylGx43/1uucKHqU76WoX6PD0JuVhauezGpw47/1OicKsxrc+G+9zonCrAY3/luvcxuFWQ1u/Lde50RhTVPm8aPc+G+9zInCmsJdSM3Ky20U1hQeP8qsvJworCnchdSsvJworCnchdSsvJwoClDGcYPchdSsvNyY3WZlbfR1F1Kz8io0UUg6Bfg0MA64KCIWD5t/OPBl4HnAByPi4+2PsrnKNm5QmbvEmlmmsEQhaRzwOeBkYCOwWtKKiLipYrF7gXcBc9of4cjGcnIsU6NvWUtHZvZ4RbZRHAusj4jbIuIR4JvA7MoFIuLuiFgN7CwiwGqGTo6btu8geOzkWG87Q5kafd0l1npNGdsX61FkohgANlR83pimdbSxnhzL1OhbptKRWS1jvUjsZkUmClWZFg2vTJonaY2kNVu3bh1DWPnGenKcc8wAi+ZOZ6C/DwED/X0smju9K6tqylQ6Mqull0vQRTZmbwQmVXyeCGxudGURsQRYAjA4ONhwwqllQn8fm6okhdGcHMsybtD8WYc9ro0Curd0ZFZLL5egiyxRrAamSZoqaS/gtcCKAuOpS5mqjsaqTKUjs1p6uQRdWIkiInZJeiewkqx77Jci4kZJZ6X5F0o6GFgDPAXYLekc4IiIeKCouH2/wOOVpXRkVksvl6AV0bJamsIMDg7GmjVrig7DzEqmzPcNSbomIgarzfOd2WZmderVErTHejIzs1xOFGZmlstVT12ozPWkZtZ5nCi6jMdXMrN2c9VTl+nlu0PNrBhOFF2ml+8ONbNiOFF0mV6+O9TMiuFE0WU8hIiZtZsbs7uMhxBpDvccM6ufE0UX6tW7Q5vFPcfMRsdVT9Zz3HPMbHRcoigZV6nU5p5jZqPjEkWJ9PKjGkfDPcfMRseJokRcpVIf9xwzGx1XPZWIq1Tq455jZqPjRFEizXiedzN0QzuJe46Z1c9VTyXSCVUqbicxKx8nihKZc8wAi+ZOZ6C/DwED/X0smju9rVfObicxKx9XPZVM0VUqbicxKx+XKKyp3PXUrHycKLrY8rWbmLl4FVMXXMbMxas6oh2gE9pJzKy5XPXUpTp1vCJ3PTUrHyeKLpXXaFz0SbnodhIzay5XPXUpNxqbWbs4UXQpNxqbWbs4UXQpNxpbK3RiBwkrntsoupQbja3ZOrWDhBWvrkQhaSpwJBDAzRFxW0ujsrq40diaqZM7SFixchOFpKcAFwGDwG8BATMkXQO8JSIeaHmEZtYW7iBhI6nVRvEZ4Cbg2RExNyL+ATgUuB74bKuDM7P2cQcJG0mtRDEzIhZGxO6hCZH5CHB8a0Mzs3ZyBwkbSa1EoVZuXNIpktZJWi9pQZX5kvSZNP86Sc9rZTxmvawTRh+2zlSrMftXkj4MfDQiYmiipA8BV41lw5LGAZ8DTgY2AqslrYiImyoWOxWYll4vAC5I/5pZC7iDhFVTq0RxNjAdWC/pu5K+I+k2YAbwzjFu+1hgfUTcFhGPAN8EZg9bZjbwlVTddRXQL+mQMW7XzMxGIbdEkXo1vUrSocARZFVR74+IW5uw7QFgQ8XnjTyxtFBtmQFgy/CVSZoHzAOYPHlyE8IzMzOo4z4KSZOBbRFxqaQpwKCkvoi4YYzbrtb+EQ0sk02MWAIsARgcHKy6TBl0w/OozaxccqueUgPzz4GrJL0V+CFZu8G3JP3PMW57IzCp4vNEYHMDy/QMP4/azIpQq43iDLIqp5nAJ4EXR8RbyNoX3jzGba8GpkmaKmkv4LXAimHLrADemHo/HQfcHxFPqHbqFX4etZkVoVbV06MRsUPSI8AOYBtARPxZGlvP2YjYJemdwEpgHPCliLhR0llp/oXA5cBpwHrgYeBNY9pol/Ods2ZWhFqJ4lpJlwBPBq4Alkr6IfAysju2xyQiLidLBpXTLqx4H8A7xrqdspjQ38emKknBd86aWSvVqnp6K3Ap8A2yaqgLyO7IXkePX90XwXfOmlkRanWP3UWWJIZcmV5WAA8tbmZFaPh5FJJ+EBGnNjMYq813zppZu9UaZnyksZUEHN30aMxszHyvjTVbrRLFarL7KKp1cepvejRmNiZ+Sp21Qq1EcTPwPyLiluEzJG2osryZFchPqbNWqNXraWHOMmc3NxQzGyvfa2OtUKtE8UxgsqSXtyMYMxsb32tjrVCrRLEvsB/ZM7P/mWzk1gHgLLKhPcysg/heG2uFWvdRnAcg6UfA8yLiwfR5IfDvLY+uwzSrN4l7pVir+F6b3tTqc0q991FMBh6p+PwIMKVpUXSBZvUmca8Ua7VuuNfGF0vN045zSq2qpyFfBa6WtFDSvwK/AZY2JYIu0ayRWz0CrPU6D5ffXO04p9SVKCLiY2RjO90HbAfeFBGLmhZFF2hWbxL3SrFe54ul5mrHOaXuITwi4lrg2qZtucs0qzeJe6VYr/PFUnO145xSb9VTz2tWbxL3Smm95Ws3MXPxKqYuuIyZi1e5SqPDjHQC88VSY9pxTnGiqNOcYwZYNHc6A/19CBjo72PR3Omjbixq1nqsOtd/dz5fLDVXO84pyp4NVC6Dg4OxZs2aosOwAsxcvKpqMXygv49fLXhZARFZNe711HkkXRMRg9XmNTzMuFkncv13d+iGLrz2GFc9Wam4/tus+ZworFRc/22dokydKlz1ZKXiISysE5RtBAYnihZxY11xOqH+2/u/t5XtuSBOFC1QtqsJGx3vfytbpwq3UbSAhyjobd7/VrZOFU4ULVC2qwkbHe9/K1unCieKFijb1YSNjve/lW0EBrdRtMD8WYc9ro4auvtqwkbH+9+gMzpVNIsTRQv0ShdN9+yprlf2v/UOj/XUJTrtpDy8Zw9kV83dXLw262Ue66nLdWJ3y7L1Ey+TTruosO7nxuwu0IndLd2zpzN5mHVrhUIShaQDJP1Y0i3p3/1HWO5Lku6WdEO7Y+wk9ZyU2z2ujHv2dKZOvKiw7ldUiWIBcEVETAOuSJ+ruRg4pV1BdapaJ+UiriLL1k+8LFzSs1YoKlHMBpam90uBOdUWiohfAPe2KaaOVeukXMRVZNn6iZeFS3rWCkU1Zj8jIrYARMQWSU8f6wolzQPmAUyePHmsq+sotbpbFnUVWaZ+4mXhezisFVqWKCT9BDi4yqwPtmJ7EbEEWAJZ99hWbKNIeSflCf19VR//6avI3uN7OKwVWpYoIuKkkeZJukvSIak0cQhwd6vi6AW+irRKLulZsxXVRrECODO9PxP4XkFxlILbC8yslQq5M1vS04BvA5OBO4BXRcS9kiYAF0XEaWm5bwAnAAcCdwH/GhFfrLX+Mt6ZbWbWSh13Z3ZEbANOrDJ9M3BaxefXtTMuMzN7It+ZbWZmuZwozMwslwcFHAUPtmZmvciJok6dOIKrmVk7OFHUqd5htWuVOlwqMbNu40RRp3pHcB2p1AFw3qU3ct/DO//22aUSs8b4gqu9nCjqVM8wGSOVOs679Eb+snP3E+YNzffDfmyIT4C1uRq4/dzrqU71DKs9Uqnjvod3Vk0Stb5XS7ufQWGt5YcO1cfP3Gg/J4o61TNMRqOD8DXyPZ9UyscnwPr4mRvt56qnUag12NpIg/M9ac892L5jZ9XvNDp4n59ZXT4+AdbHoyW3n0sUTTRSqWPhK498QrUVQH/f+IYH7/NJpXz80KH6+OmK7ecSRZPllTqa2Ujpq6ry8XDx9fEzN9qvkNFjW60XRo8d3vMDspOKhxfvbu71ZEXpuNFjbex64aqqF0+afuiQdSInii5W5pOK+8qbdQ43ZltHcldRs87hRGEdyb26zDqHE4V1JHcVNescThTWkdxX3qxzuDHbOlIv9Ooy6xZOFNaxytyry6ybuOrJzMxyOVGYmVkuJwozM8vlRGFmZrmcKMzMLJcThZmZ5XKiMDOzXE4UZmaWy4nCzMxyOVGYmVmuQhKFpAMk/VjSLenf/assM0nSTyXdLOlGSe8uIlYzs15XVIliAXBFREwDrkifh9sF/EtEPAc4DniHpCPaGKOZmVFcopgNLE3vlwJzhi8QEVsi4tr0/kHgZsAjxJmZtVlRieIZEbEFsoQAPD1vYUlTgGOA3+QsM0/SGklrtm7d2sxYzcx6WsuGGZf0E+DgKrM+OMr17At8FzgnIh4YabmIWAIsARgcHIzRbMPMzEbWskQRESeNNE/SXZIOiYgtkg4B7h5hufFkSeLrEbGsRaGamVmOoqqeVgBnpvdnAt8bvoAkAV8Ebo6IT7QxNjMzq1BUolgMnCzpFuDk9BlJEyRdnpaZCZwBvEzSb9PrtGLCNTPrXYU8CjUitgEnVpm+GTgtvf8loDaHZmZmw/jObDMzy+VEYWZmuZwozMwslxOFmZnlcqIwM7NcThRmZpbLicLMzHI5UZiZWS4nCjMzy+VEYWZmuZwozMwslxOFmZnlcqIwM7NcThRmZparkGHGbWTL127i/JXr2Lx9BxP6+5g/6zDmHDNQdFhm1sOcKDrI8rWbOHfZ9ezY+SgAm7bv4Nxl1wM4WZhZYVz11EHOX7nub0liyI6dj3L+ynUFRWRm5kTRUTZv3zGq6WZm7eBE0UEm9PeNarqZWTs4UXSQ+bMOo2/8uMdN6xs/jvmzDisoIjMzN2Z3lKEGa/d6MrNO4kTRYeYcM+DEYGYdxVVPZmaWy4nCzMxyOVGYmVkuJwozM8vlRGFmZrkUEUXH0HSStgJ/atHqDwTuadG6G+WY6teJcTmm+nViXGWJ6ZkRcVC1GaVMFK0kaU1EDBYdRyXHVL9OjMsx1a8T4+qFmFz1ZGZmuZwozMwslxPF6C0pOoAqHFP9OjEux1S/Toyr9DG5jcLMzHK5RGFmZrmcKMzMLJcTRQ2SXiXpRkm7JeV2N5M0TtJaSd8vOiZJkyT9VNLNadl3Fx1TWu4USeskrZe0oJUxpe0dIOnHkm5J/+4/wnLvSfHfIOkbkvbugJj6JX1H0u/Tfjy+6JjSsu06zmvG1K7jvNZxq8xn0vzrJD2vFXE0ENfpKZ7rJF0paUYj23GiqO0GYC7wizqWfTdwc2vDAeqLaRfwLxHxHOA44B2SjigyJknjgM8BpwJHAK9rcUwAC4ArImIacEX6PDyuAeBdwGBEPBcYB7y2yJiSTwM/jIjDgRm09tiqNyZo33FeT0wtP87rPG5PBaal1zzggmbGMIa4bgdeEhFHAR+lwUZuJ4oaIuLmiFhXazlJE4GXAxd1QkwRsSUirk3vHyT7j92yB13U+TsdC6yPiNsi4hHgm8DsVsWUzAaWpvdLgTkjLLcn0CdpT2AfYHORMUl6CvB3wBcBIuKRiNheZEwprrYd5/XE1KbjvJ7jdjbwlchcBfRLOqTJcYw6roi4MiLuSx+vAiY2siEniub5FPA+YHfBcTyBpCnAMcBvCg5lANhQ8XkjLUxeyTMiYgtkJxXg6cMXiIhNwMeBO4AtwP0R8aMiYwKeBWwFvpyqeS6S9OSCY4L2Huf1xgS09Div57gt4tge7TbfAvygkQ35CXeApJ8AB1eZ9cGI+F4d3/974O6IuEbSCZ0QU8V69gW+C5wTEQ8UHJOqTBtz/+y8uOr8/v5kV2JTge3Av0t6Q0R8raiYyP5vPg84OyJ+I+nTZFUvHyoqpnYf56NcT9OO82qrrzJt+HHbkmO7hrq3KemlZIniRY1syIkCiIiTxriKmcArJZ0G7A08RdLXIuINBcaEpPFk/3m+HhHLxrq+JsS0EZhU8XkiTajiyYtL0l2SDomILakq4O4qi50E3B4RW9N3lgEvBBpOFE2IaSOwMSKGro6/Q367QTtiautxXmdMTT/Oq6jnuG3Jsd2EuJB0FFlV4akRsa2RDbnqqQki4tyImBgRU8gaQVeN5T9PM0gSWf32zRHxiSJjqbAamCZpqqS9yH6rFS3e5grgzPT+TKBayecO4DhJ+6Tf7URa21hbM6aIuBPYIOmwNOlE4KaCY2r3cV4zpjYd5/UctyuAN6beT8eRVV9uaVE8dcclaTKwDDgjIv7Q8JYiwq+cF/APZJn7r8BdwMo0fQJweZXlTwC+X3RMZEXMAK4DfptepxX9OwGnAX8AbiWrsmr1/nsaWY+ZW9K/B4wQ13nA78l6b30VeFIHxHQ0sCbtw+XA/kXHVLF8O47zmjG16zivdtwCZwFnpfci64F0K3A9WQ+6lh7bdcZ1EXBfxW+zppHteAgPMzPL5aonMzPL5URhZma5nCjMzCyXE4WZmeVyojAzs1xOFGYNSqO7vr3oOMxazYnCrHH9gBOFlZ4ThVnjFgOHSvqtpPMlzZe0Oo39fx5kA9Upe57ERcqedfF1SSdJ+lV6zsKxabmFkr4qaVWa/rZC/zKzCk4UZo1bANwaEUcDPyZ7FsGxZHdUP1/S36Xlnk32bImjgMOB15PdUfxe4AMV6zuKbAjv44EPS5rQ+j/BrDYnCrPm+G/ptRa4liwhTEvzbo+I6yNiN3Aj2cN4gmyohykV6/heROyIiHuAn5IlHbPCefRYs+YQsCgivvC4idkzEv5aMWl3xefdPP7/4PDxdDy+jnUElyjMGvcgsF96vxJ4c3ouApIGJOU+aKeK2ZL2lvQ0skH3VjctUrMxcInCrEERsS01St9A9uSwS4BfZyNf8xDwBuDRUazyauAyYDLw0Yho9fMMzOri0WPNOoCkhcBDEfHxomMxG85VT2ZmlsslCjMzy+UShZmZ5XKiMDOzXE4UZmaWy4nCzMxyOVGYmVmu/w+LyNOAe3JHXwAAAABJRU5ErkJggg=="
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Writing a function that can do this for any of the organisms on file:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "source": [
    "def exploration(d18O_raw_filename, temp_raw_filename):\n",
    "    paleo_data = pd.read_csv(\"d18O_raw/\" + d18O_raw_filename)\n",
    "    temp_data = pd.read_csv(\"temp_raw/\" + temp_raw_filename)\n",
    "    \n",
    "#     print(paleo_data)\n",
    "    \n",
    "    temp_data['year'] = temp_data['time'].apply(lambda x: x[0:4])\n",
    "    \n",
    "#     print(temp_data)\n",
    "    \n",
    "    actual_annual_temp = temp_data.groupby('year')['temp'].agg(np.mean).to_frame().reset_index()\n",
    "    actual_annual_temp['year'] = actual_annual_temp['year'].apply(lambda x: int(x))\n",
    "    \n",
    "    if d18O_raw_filename == 'amchitka_d18O_raw.csv':\n",
    "        paleo_data['year'] = paleo_data['Year']\n",
    "        paleo_data['d18O'] = paleo_data['AM-KR-80_d18O']\n",
    "        del paleo_data['Year']\n",
    "        del paleo_data['AM-KR-80_d18O']\n",
    "    \n",
    "    temp_and_algae = pd.merge(actual_annual_temp, paleo_data, on='year')\n",
    "    temp_and_algae.set_index('year')\n",
    "    \n",
    "    print('pearson correlation:')\n",
    "    print(np.corrcoef(temp_and_algae['temp'], temp_and_algae['d18O']))\n",
    "    \n",
    "    plt.plot(temp_and_algae['year'], temp_and_algae['temp'])\n",
    "    plt.plot(temp_and_algae['year'], temp_and_algae['d18O'])\n",
    "    plt.title(d18O_raw_filename.split(\"_\")[0] +\": Temperature and D18O\")\n",
    "    plt.xlabel('Year')\n",
    "    plt.legend(['temp', 'd18O'], loc='upper right')\n",
    "    plt.show()"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "source": [
    "exploration('amchitka_d18O_raw.csv', 'amchitka_SODA_1x1_raw.csv')"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "pearson correlation:\n",
      "[[ 1.         -0.27061659]\n",
      " [-0.27061659  1.        ]]\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "source": [
    "exploration('attu_d18O_raw.csv', 'attu_SODA_1x1_raw.csv')"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "pearson correlation:\n",
      "[[ 1.         -0.42953347]\n",
      " [-0.42953347  1.        ]]\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "source": [
    "exploration('kingitok_d18O_raw.csv', 'Kingitok_SODA_1x1_raw.csv')"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "pearson correlation:\n",
      "[[ 1.         -0.00540379]\n",
      " [-0.00540379  1.        ]]\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "source": [
    "#exploration('maine_d18O_raw.csv', 'maine_SODA_1x1_raw.csv')"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "source": [
    "exploration('qikA_d18O_raw.csv', 'qikA_SODA_1x1_raw.csv')"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "pearson correlation:\n",
      "[[ 1.         -0.29826873]\n",
      " [-0.29826873  1.        ]]\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAEWCAYAAABliCz2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAABSGUlEQVR4nO2dd3hc1bW33zWjbkmWVaxiyZZ7t+Um24BpNthgTA09AQIESEiAfEkuEAIhCcnl3psGIYQSOpgWOhhsDBhscO9d7rJsdVm9a/b3x56xx7LKjKZK2u/zzDNnztnnnDVnpN/ZZ6211xalFAaDwWDo+VgCbYDBYDAY/IMRfIPBYOglGME3GAyGXoIRfIPBYOglGME3GAyGXoIRfIPBYOglGMHvxYjILBHZ7fT5oIjMCaRNBt8jImeLSF6g7TD4HyP4vRil1HKl1Eh39hGRZSJyTETCXWw/UESqnV5KRGqcPs/qmvWBxf49hgXaDl/Q6jcqFZEvROTqVm2uEpHvRKRWRJa1cYxzRWSDiFSKyH4Rua3V9jEi8qGIVIhIlYh8JSKn+fir9XqM4BtcRkQygVmAAi52ZR+lVK5SKtrxsq+e6LRuuY/M7TIiEuKHc1h9fQ4PmWj/vUYCLwJPiMhvnbaXAX8HHm29o4iEAu8BTwN9gauBv4rIRPv2ocC3wFZgMJBmb79ERGb66PsYMILf4xGRSfaeVpWIvCkib4jII/Zt7T7ai8goETkgItc4rb4BWIUWgBu9YFu4iPxZRHJFpFBEnhKRSGfbROS/RKRIRPJF5FIRuVBEckSkTER+7XSsh0XkP/bvWGX/zhOdtqeJyDsiUmz/Xne1se+rIlIJ3CQi2SKyUkTK7ed+QkTC7O2/se+62d4LvlpEbhKRFa2+3/GnABF5UUT+JSKLRKQGOKcjm9q4VvNFZKO9x3xYRB522pZpP9eN9mtZIiIPOG2PtJ//mIjsAKa5+hsppUqUUq8APwbuF5EE+/qlSqm3gKNt7BYPxAKvKM1aYCcwxr79YWClUuoBpVSZUqpKKfU48ArwP67aZugCSinz6qEvIAw4BPwcCAW+BzQBj9i3nw3kObU/CMwBJgO5wEWtjrcX+AkwxX6cZKdtTwJPumCTAobZl/8OfIgWiBjgI+C/nWxrBh6y2/4joBhYaG87FqgHhtjbP2y36Xv29r8EDtiXLcB6+7HCgCHAfmBuq30vtbeNtH/HGUAIkIkWrHva+h72zzcBKzr4ri8CFcDp9nNEdWRTG9ftbGC8fd8JQCFwqX1bpv1cz9ptnwg0AKPt2x8Fltuvcwawzfl37+g3cloXav89Lmi1/lZgWRvHWAjcCViBmUARkGHfVgD8sI19zgFagKhA/+/01FfADTAvH/64cCa6ByZO676jY8H/HZAHnNPqWGfYRTHR/nkX8PMu2KSAYYAANcBQp20zgQNOttUBVvvnGPu+053ar3cSvYeBVU7bLEA+2gU1HchtZcf9wAtO+37Tid33AO+1/h5On2+ic8F/2Wlbhza5cB3/DvzNvpxpP1e60/Y1wDX25f3APKdtt7kr+Pb1BcD1rda1J/gL0DelZvvrR07bmp3tcVo/yn7uAYH8v+nJL5/7Kg0BJQ04ouz/TXYOdbLPHcDXSqmvWq2/EViilCqxf15oX/e3LtqWhL2XKyKOdYLuETooVUq12Jfr7O+FTtvrgGinz4cdC0opm91dlYYWkTQRKXdqa0X3ek/ZF0BERgB/Baba7QxB32A8wfkcg1ywydme6eie+jj0E0E48HarZgVOy7WcuDZprc7d2d9AW+cPRf9mZS60HQW8CVwGfA4MBz4WkaNKqU+AEiC1jV1TARtwzF37DK5hfPg9m3xggDgpKjCwk33uAAaKyHEht/vVrwLOEpECESlAu4kmOvvJ3aQELdhjlVJx9ldfdSKw2xUynGy2AOnoJ5zD6CeHOKdXjFLqQqd9W5eN/Rf6KWa4UioW+DX6htQeNegbg+P8KW20cT6HKzY5sxDt/spQSvUFnurEHmfycbo2dP430BaXoHvma1xoOw7YrZRarJSyKaV2A58AF9i3LwWubGO/q9C+/dou2GdwASP4PZuV6H/Su0QkREQuB7I72acKmAecKSKODIxL0b7VMUCW/TUa3Ru9oSuGKaVsaJ/z30SkP4CIDBCRuV05np0pInK5PcvmHrQfexVapCpF5F57ANMqIuNEpKPgZQxQCVTbe6w/brW9EO13d7AZGCsiWSISgXYTdYS7NsUAZUqpehHJBq7r5PjOvIUOuPYTkXTgZ67uKCLxInI98E/gf5RSpfb1Vvv3DAEsIhJhfwoA2AgMF52aKaKzci5CXyPQbsPTROSP9uPHiMjP0H9L97rxvQxuYgS/B6OUagQuR/uXj6HT4951Yb9y4DzgAhH5A9p184LSKZYFjhfwBHC9/WbylIg85aaJ96IDwavs2TFL0WmAXeUD9Hc8BvwAuFwp1WR3Cy1A36gOoJ8u/o1OGWyPX6JFtQp9Y3qz1faHgZfsWTxXKaVygN/bv8MeYAUd0AWbfgL8XkSq0IHetzo6fit+h3bjHACWoLNhOmOziFSjf59b0fGah5y2/wD9hPYvdJykDn2dUErtA24GHkffNL8G3gGes2/fg44JTUTHjfKBK9AB62/d+F4GN5GT3buGno6IvIgO2P0m0LZ4E3ua4jCl1PcDbYvBEKyYHr7BYDD0EozgGwwGQy/BuHQMBoOhl2B6+AaDwdBLCOqBV4mJiSozMzPQZhgMBkO3Yf369SVKqaS2tgW14GdmZrJu3bpAm2EwGAzdBhFpdyS1cekYDAZDL8EIvsFgMPQSjOAbDAZDLyGoffgGg8HgCk1NTeTl5VFfXx9oU/xGREQE6enphIaGdt7YjhF8g8HQ7cnLyyMmJobMzExOLg7bM1FKUVpaSl5eHoMHD3Z5P+PSMRgM3Z76+noSEhJ6hdgDiAgJCQluP9EYwTcYDD2C3iL2DrryfY3gGwwGg5doarZRVtOIzRacJWuM4BsMBoOHlJeX8+STT1JS00DesVpyCqsor20k2GqVGcE3GAwGD3EIfmOzjRCrBatFyC2rZV9xDbUNzYE27zhG8A0Gg8FD7rvvPvbt28e8s2by9z8+xHsv/YsbL5nD/LNn8P/ue4Dcslpy9u5n1KhR3HrrrYwbN47rr7+epUuXcvrppzN8+HDWrNHTBT/88MP84Ac/4Nxzz2X48OE8++yzXrPTpGUaDIYexe8+2s6Oo5VePeaYtFh+u2Bsu9sfffRRtm3bxpuLl7Nl9Td89dlHbFi3luYWGxdedBFffLWMtAHp7N27lzfefItnnnmGadOmsXDhQlasWMGHH37In/70J95//30AtmzZwqpVq6ipqWHSpEnMnz+ftLQ0j7+H6eEbDAaDl2hRihXLvmTJkiVMmjSJaVOncGDvHhpLjxAdHsqAjEGE98+kodnG2LFjmT17NiLC+PHjOXjw4PHjXHLJJURGRpKYmMg555xzvPfvKaaHbzAYehQd9cR9iSMxxyJw//33c/vtt5+0/eDBg/SJisSmFEVVDVgsFsLDw/U+FgvNzSd8/a1TLr2Vcmp6+AaDweAhMTExVFVVATB37vk8//zzVFdXA3DkyBGKiooAfTOIiwqjsr65wwyeDz74gPr6ekpLS1m2bBnTpk3zip2mh28wGAwekpCQwLTpM7h89kwuXTCf6667jpkzZwIQHR3Nq6++itVqBSAuMpTS6gYaW9oX/OzsbObPn09ubi4PPvigV/z3YATfYDAYvMLjT79AVUMzo1NjAbj77rtPabNt2zaUUoSFWPjj355kSFI0oCd72rZt2/F2I0aM4JlnnvG6jcalYzAYDF6gocVGmLVzSRUR4iLDqG5opqnF5gfLTmB6+AaDoV12F1RRXNXAGcMTA21K0NPYbCM63DVJjYsKpaiqnvLaJpJiwk/a9vDDD/vAOo0RfIPBcAq1jc38fekenltxgBab4u9XZ3HppAGBNitosSlFU4uNsBDXnCYRoVYiw6yU1zaeIvi+xAi+wWA4iS92FvLQB9s5Ul7HNdMyOFRayy/f3kxsZAjnjkoOtHlBSVOzds244tJxEBcZRn5FHfVNLUSEWn1l2kkYH77BYACgoKKeH7+6nlteWkdUmJW375jJo1dM4JkbpjA6NZYfv7qBtQfLAm1mUNJo98W72sMH7dYRoLy2yUdWnYoRfIOhl9NiU7z03UHm/PVrvtxVxK/mjuSTu2YxLTMegJiIUF784TQGxEVy84tr2Znv3bIFPYGGZvcFP9RqoU94COV1/quq6ZHgi0i8iHwuInvs7/3aaXdQRLaKyCYRWefJOQ0Gg/fYfrSCy5/8lt9+uJ1JA+NYfM+Z3HnOsFOEKyE6nFdunU50eAg3PL+GQ6U1AbI4OGlstmERIcSiR8Q+/PDD/PnPf+btt99m7NixWCwW1q07IX1NTU3ceOONXHT2DC6cNY3fP/Kn49sqKiq44YYbGDp0KEOHDuWGG26goqLCK3Z62sO/D/hCKTUc+ML+uT3OUUplKaWmenhOg8HgBSpqm7jyqZUcKa/jsWuyePnmbDIT+7TbfkBcJK/ckk1zi40fPLeGosreM2F4ZzQ265TM1iUQxo0bx7vvvsuZZ5550vq3336bhoYGtm7dwpufLuPfzz5zvJbOLbfcwpAhQ9i3bx/79u1j8ODB3HrrrV6x01PBvwR4yb78EnCph8czGAx+4rt9JdQ2tvDk9VO4JGuAS/VahvWP4YUfZlNS3cANz6+hwo/+52CmscXGM4//mZEjRzJnzhx2794NwOjRoxk5cuQp7UWEmpoalM1GmGrCGhJKdEwMe/fuZf369Tz44IPH2z700EOsW7eOffv2eWynp1k6yUqpfAClVL6I9G+nnQKWiIgCnlZKtTuETERuA24DGDhwoIfmGQyG9li+t4To8BAmDYxza7+sjDie+cFUfvjiGm55aS2v3DKdyDD/ZJm4xKf3QcFW7x4zZTxc8Gibm5RSbNywgU/ef4eNGzfS3NzM5MmTmTJlSruH+973vscHH3xAamoqtbW1/OKhPxIWFcuOb78lKyvreBkGAKvVSlZWFtu3b2fo0KEefY1Oe/gislREtrXxusSN85yulJoMXADcKSJnttdQKfWMUmqqUmpqUlKSG6cwGAzusGJPCTOGxBPqRiqhgzOGJ/LYNZNYn3uMn7y23u8jRoOJZpti/ervmH/xxURFRREbG8vFF1/c4T5r1qzBarVy9OhR9u3fzyvP/JMtO3JQSrX5pNXeenfptIevlJrT3jYRKRSRVHvvPhUoaucYR+3vRSLyHpANfNNFmw0Gg4fkltaSW1bLzadndvkYF45P5ZFLx/HAe9t4f+MRrpya4T0DPaGdnrivaLRn6IRYXL9xLly4kHnz5hEaGkpKcjLTZ8xk1dq1LDj3dDZu3IjNZsNiP57NZmPz5s2MHj3aY1s99eF/CNxoX74R+KB1AxHpIyIxjmXgfGBb63YGg8F/LN9bDMAZwz17ir4ueyAJfcJYub/UG2Z1S5pabEyZfhqffPQhdXV1VFVV8dFHH3W4z8CBA/nyyy9RSlFTU8OmDevIHDqM/gMGMWnSJB555JHjbR955BEmT57MsGHDPLbVU8F/FDhPRPYA59k/IyJpIrLI3iYZWCEim4E1wCdKqc88PK/BYPCAFXtKSO0bwdCk9rNyXEFEyB4cz5oDvXdAVkOzjdHjJ3L11VeRlZXFFVdcwaxZswB47733SE9PZ+XKlcyfP5+5c+cCcOedd1JdXc24ceOYNm0aN//wJsaNm0B5XRPPPfccOTk5DBs2jKFDh5KTk8Nzzz3nFVs9CtoqpUqB2W2sPwpcaF/eD0z05DwGg8F7tNgU3+0r5fwxyV7xC2cPjufTbQUcKa9jQFykFyw8mRabYt3BMtLiIsmIj/L68T2lsdlGqNXCb37zG37zm9+csv2yyy47ZV10dDRvv/32SesKKuoormogPbUvr776qk9sNbV0DIZextYjFVTUNXmtAmb2YD0id+2BMgZ4qcCazaZYe7CMj7fk8+m2AkqqG5gxJJ43bpvpleN7k0YXyyJ3RlxUGEVVDVTUNZEY7ZuCakbwu8iba3NZsbeUf1w7KdCmGAxusWKP9t+fPsw7gj8qJZaYiBBWHyjzqKKmzabYePgYH23OZ9HWfIqqGogItXDuqP6U1TSy42il17JVvIk7ZZE7IiLUSmSolfJaI/hBxydbC/gmp5j/vny8V35sg8FfLN9TwpjUWK+JitUiTMuMZ82BrgVum1ps/O3zHN7feISjFfWEhVg4e0QSF01MY/ao/vQJD+HllQdZtb+Mgsp6Uvu27TYKxM3AZnOvLHJnxEXpCpoNTS2Ed1JBsyv1d4xSdZGcAj1h8e6CSqYMig+wNQaDa9Q0NLMh9xg3nz7Yq8fNHhzPl7uKKKlucPtGsmhrPk8u28dZI5L41byRzBmdTExE6EltRibHALCroKpNwY+IiKC0tJSEhAS/in5XqmR2RFxkKPkVdZTXNZHcgeArpSgtLSUiIsKt4xvB7wIVdU0U2OuI7DhqBN/QfVhzoIymFuX1Gayc/fgXjE91a98vdhYR3yeM52+ahtXStliPTNGCn1NQxTkjTx3Qn56eTl5eHsXFxW5a7hn1TS2UVDeijoVT4CXRr6hqoDRPUda3YzGPiIggPT3drWMbwe8Cewqrji/vyK/qoKXBEFws31NCWIjleOljbzEurS+RoVZWuyn4TS02lu0u4rwxKe2KPWhXR3JsOLsL2v5/Cw0NZfBg7z61uMJL3x3ktx9uZ80Ds+kf415vuz2+2FnIgZIazhiR2aVR0B1hBL8L5BRWAzAwPoodpja4oRuxYm8x2ZnxXp9hKSzEwuRBcW7n4687eIzK+mbmjG6vDNcJRqbEsqsdwQ8UuWW1RIZaSfJikHX2aN/NKmYmQOkCOYVVRIVZmT26P7sLKmmx+WfyAoPBEwor68kprPbZhOTZmQnsLKikos71Cppf7CwkzGph1ojOR/yOSolhb3E1zUFUtye3rJaB8VFBlznUHkbwu0BOYRXDk2MYm9aX+iYbB0rMZBCG4GfFnhIAzvBSOmZrsgfHoxSsP+R6L/+LXUVMHxLvUqbbyOQYGpttHCyt9cRMr3K4rDYoB4O1hxH8LpBTWM2I/tGMSY0FMFO+GboFK/aWkNAn7PjfrbeZNDCOUKuw2kW3zr7iag6U1DDHRReGI3Dbnh/f3yiljvfwuwtG8N2krKaRkuoGRqbEMKx/NKFWMX58Q9CjlGLF3hJOG5aIpYPgqCdEhFqZmB7H6v2uCf4XOwsBmO2C/x5gWP9oLKJToYOBkupGahtbGBjv/XISvsIIvpvk2DN0hifHEBZiYVj/mG7fw7/3P1t48dsDgTbD4EN2F1ZRXNXALB+5cxxkD45n25EKahqaO227dGcRo1JiSO/nWg85ItRKZmKfoAnc5pZp19KgBM8K0PkTI/hu4hD8EcnRAIxOjWHH0e4r+EfK63hz3WEe+2IP9U0tgTbH4COO++99FLB1MH1IAs02xcbc8g7bldc2sv7QMZd79w5GpcSwuzA4BP+wXfCND78Hk1NYRUxECCmxOud2TGosRVUNlFQ3BNiyrrFkewEAx2qb+HRbfoCtMfiK5XtKGJLUhzQfVLN0ZsqgfliETsssfJ1TTItNuZ2CODI5ltyyWmobO3+C8DWOHn56P+PS6bHkFFQzIjnmeBpWdw/cfratgGH9oxmS2IdXV+UG2hyDD2hobmH1gVKfu3MAosNDGDegb6eB26U7i0iMDiMrPc6t449MiUYp2GMfCxNIcstqSYmN8PqYBl9iBN8NlFLkFFUxwl7XA2B0Nxb80uoG1h4s44JxKVw3fSDrDx3rlt/D0DHrDx2jvsnm8exWrpKdGc/Gw+U0NLftInSMrj1nZH+3A8gjU/T/WzBk6uSWdq8MHTCC7xbFVQ2U1zYd998D9OsTRmrfiG7px1+6sxCbgrljU/jelHTCQyy8uupQoM0yeJkVe0qwWoQZQ/xT8yl7cDyNzTa25FW0uX3twTKq6pu7NKJ0YHwUEaGWoAjc5pbVMjDBCH6PxVFSYaRTDx90L39nN6yp89m2AtL7RTI2LZa4qDAumpDG+xuPUO1ChoWh+7BibwmTMuJOqUDpKxx1etors/DFziI9urYLAWSrRRiRHHM8eSJQ1De1UFBZb3r4PZndTimZzoxJjWVvcXW3ynKprG/i272lzBubcjwe8f0ZA6lpbOH9jUcCbJ3BWxyraWTrkQqfZ+c4069PGCOTY9r04yul+GJnITOHJtCni/NIjEyOCXgPP+9YHYAR/J7MnsIq4vuEkRgddtL60amxtNgUe4sCH0hyla92FdHYYmPuuJTj67Iy4hibFsurqw51aXIFQ/Dx3b5SlKJLvWlPyB4cz/qDZafUvdlXXMPB0lqXiqW1x8iUGEqqGygNYGZcd0zJBCP4brG7sIrh/aNPKZQ0Jk0HkrqTH3/x9gISo8OZPLDf8XUiwvXTB7GroIoNuccCaF33pKKuieV7inniyz389oNt1DX69olv25EKdhVUdnhzXrG3mJjwECa6mQ3jKdmD46lpbDllFLpjdO25HlSEDIYSC4dKdf2s7tbDN+WRXUQpxd7Cai6bfOqcnYPio4gKs3abEgv1TS0s213MpZMGnFKD/JKsNP60aCevrco1E7t0QENzCzvzq9h8uJzNh8vZlFfO/uKTi+hNzIjj8snuTVDhzvmvfGoldU0tpPeLZPao/swZk8z0wQnHZ19SSrF8TwkzhiYQ4uW66p3hmBBlzYEyJjjdbL7YWcTo1FgGeDAe4LjgF1Zxmh9STdsit6yOqDDrKU/7wY4RfBfJr6inqqH5FP89gMUijErpPiUWlu8pobaxhXljU07Z1ic8hMsnD+CNtYd58KIx9OvTvf6g/cH/fLaL55YfOD69XVJMOFkZcVwxOZ2J6XGMH9CXCx9fzkebj/pM8LcfraSuqYVrswdSVFnPG2sP89LKQ0SHh3DmiERmj0pmSFIf8o7VcduZQ3xiQ0ckx0aQmRDFqv1l3DpLn/9YTSPrDpVx5znDPDp2UnQ48X3CAtrD725lkR0YwXcRR8C2dYaOg9GpsXy4+WhAJlJ2l8+2FRAbEcKMIQltbr9++iBeXnmIt9cf5rYzh/rZuuCmsLKeZ77Zz5nDE7lqagZZA+NIiY045Te/aGIqzy0/wLGaRp/cNB2lC+6ZM5zk2AjqGlv4dm8JX+wq5IudRSzaWnC87Sw/5d+3JntwPIu3F2KzKSwWYVlOETbl+QQfIhLwwO3hbpiSCUbwXWZPqxo6rRmTFstrq3M5Ul7ncjGoQNDUYmPpzkLmjE5ud+LlkSkxTMvsx2urc7n1jCFeqa54oKSG11YdotmmaLEpWpTC1mrZIsJPzhnKsP5t31SDgTfXHqbFpnj44rEdFs1aMCGNp7/ez2fbC7g2e6DX7diQe4wBcZEk20t8RIZZmTMmmTljkrHZFFuPVPDFzkIaWxSZARKm7MEJvLUuj5yiKkalxLJ0ZxFJMeFMGNDX42OPTInhrXWHj99M/ImjLLI/M5+8hRF8F9ldUE1STDhxUW331hwjbnccrQxqwV+9v4yKuqaTsnPa4vszBnH3G5v4dl+JV3qIz684wKurDxEbEYrVIlgELCL2Zf1++Fgt/WMjuO+CUR6fzxe02BRvrMll1vDETiskjk2LZUhiHz7afNQngr8pt5xJA+Pa3GaxCBMz4piY0fZ2fzHdyY8/JDGab3YXc+H4VK8I9MiUGGobW8g7Vuf3nnZxdQN1TS0MMj38nsueoqp23Tmgq/iJwM78Ks5vwzceLCzeXkBEqIUzOxHxeeNSiO8TxqurDnlF8LceqSA7M543b5/Zbpu5f/vmpAnig41lu4s4WlHPQwvGdNpWRLhoYhr/+HIPRZX19I/1zgTXoN1KR8rruPkM/0/a7Q7p/SJJ7RvB6gNlDE2Kpqqh2e3qmO3hHLj1t+B315RMMGmZLmGzKfYUVjO8HXcOQFRYCIMT+rAjv+3h5MGAzaZYvL2As0f0JzKs44JP4SFWrpyaztKdRRRU1Ht03uYWGzvzKxnfyaP88ORocoqCV/AXrs4lKSbcZR/0ggmpKAWLtnq3CulGe8rs5HZ6+MGCiJA9OJ41B8pYurOQsBCL19wgjnpW7kyGUlnf1G59H3dwVMnsbimZYATfJfKO1VHX1NJhDx9gdFpwl1jYeLicoqoG5nXiznFwffYgbErxxlrPqmjuKaqmodnG+PSOBX9EcgyHy+qCovRta46U1/HV7iKumZZBqIspjsOTYxiVEsNHW7wt+OWEWS3Hx38EM9mD4ymuauA/6/M4fWgCUWHecSpEh4eQER/pcuC2ucXGRY+v4Pcf7fD43LmldYjgUWppoDCC7wLtlVRozZhUXau7qr7JH2a5zeLtBYRahXNGufZYPTAhijOHJ/HGmsOnjJh0h61H9FPPuE56+I6AeDCOWH5zTS4KuHpahlv7LZiYxvpDx8g75r2JtzfkHmPcgFjCQ4K/LK/Dj9/VYmkdMTI5xuXUzM93FJJbVsvKfR3X6XeF7lgW2YERfBdoPctVezhq4we6zkdbKKX4bFsBpw1NpG+k60W0vj9jEAWV9SzdWdTlc287UkF0uHZ5dYTjhpoTBLXOnWlqsfHG2sOcPSLJ7YD8gglpAHzipV5+U4uuQjnJaYR0MDM0KZoEe1qqt/z3DkamxHCgpMYlN83LK3UV2P0lNVTUetYhyy2r6ZbuHDCC7xI5hVWk9Y3otNqgc6ZOsLEzv4rcslqX3TkOzhmZRGrfCF5b3fWyyVuPVDAmLbbT7IxB8VGEWS1BF7j9YmcRRVUNXD99kNv7DkyIYmJGHB9tOeoVW3bmV9LQbGs3QyfYEBFmj+7P9MHxpPb1rgtkZEoszTZ1ygjn1uQUVrFyfyln2EflbjlS7tF5HYOuuiNG8F0gp7CaESmd54Ynx+oRgME44nbx9gJEYI6bj9UhVgvXTBvI8j0lFFa6H7x1NWDrONeQpD4BL33bmoVrckntG8HZI7uWrbRgQirbjlSyv9jzJxfHgKvJ3aSHD/Dfl0/gtVune/24o1ysqfPKykOEhVj442XjANh8uLzL56xvaqGwssEIfk+lucXGvqLqk2a5ag8R0ZOaB6ngTxsUT1JMuNv7njNKC1179c07QpeNtrkk+IC91nnwuHRyS2v5JqeYq6dldLkezUUT0hCBj73g1tmQe4zk2HBS+3ovzdPXWC3ik1o+gxP7EGqVDl2oVfVNvLshjwUT0hiU0IehSX3Y5IHgO2Ix3XGULXgo+CJypYhsFxGbiEztoN08EdktIntF5D5PzulvDpXV0thic0nwAUanxLK7oMqjIKe3OVBSw66Cqk4HW7XHmNRY+oRZuyT4W/McAVvXMkpGJEdzpLyOmiCZhOX1tblYLcI107o+eCqlbwTTMuOPl97whI255Uwe2C/oy3f4g1CrhaFJ0R2mZr674Qg1jS3cMFO747Iy+rHpcEWXf4fcbpyDD5738LcBlwPftNdARKzAP4ELgDHAtSLS+ciVICGnwLWArYMxabE0NNs4UNKxX9GfLN6u66rMHdu1LIkQq4XJg/p1SfC3HakgKszK4ETXrp8jcLsnCDJ1GpttvL3uMOeO6k+Khz3qBRPT2FtUfTzjqyuUVDeQW1bbbfz3/mBkSvtPhEopXl558KRRx1kZfSmpbuBoF8eWHCrVgj+oNwq+UmqnUmp3J82ygb1Kqf1KqUbgDeAST87rT3IKqxGBYf1dE6zjgdsgcut8tq2A8QP6elTyYfrgeHYXVlFe2+jWfluPVDA2LfaUMsztMeJ4pk7g/fhLdhRQUt3I9dM9L41wwbgUrBbho81dD946/PfdJUPHH4xMieFIeR2VbaRCf7evlH3FNdww40Sw3SH8m+zX0l1yy2rpE2YlvptWkfWHD38AcNjpc559XZuIyG0isk5E1hUXF/vcuM7IKaoio1+UywNGhiZFE2a1BI3g7y+uZtPhcrezc1qTPVhX1lx70PWJUZpbbOzIr+w0/96ZgfFRhIcER6bOwtW5pPeL7LQMhSskRodz2tAEPtqc32V3wsbcY4RYxOV4SG/AEbjNacOP/9J3B4nvE8b8CalO7WMJC7GwOa+8S+c7XFZLRjcsi+ygU8EXkaUisq2Nl6u99LauTLt/8UqpZ5RSU5VSU5OSAlPW1ZmcgiqX/fcAYSEWhvWPDviI24MlNdz/7hbm/X054SEW5o9P7XynDpiQ3pcwq4U1B1wfuLKvuMatgC3oAN/QpOiAB273F1fz3b5Srs0e6LVqjAsmppFbVsuWvK6V39iQe4wxabHdcsCPr3D8b7YO3B4pr2PpzkKunpZx0vUKC7EwNi22y4Hb7pySCS4IvlJqjlJqXBuvD1w8Rx7gPDwxHfBOUrKPabT74l313zsYkxYbsFz8HUcr+dnrGzn3L8t4Z8MRrpqWztL/dxaZiR0PeuqMiFArWRlxrHGjh+8YYetuj3REcnTAe/ivr8klxCJcOdV7E5jMHZtCqLVrbp1m+4Cr7pSO6Q8GxEUSEx5yigvwtVV63Ehb7riJ6XFszatwO7GivqmFQ6W1Hv8vBRJ/uHTWAsNFZLCIhAHXAB/64bwec6CkhmabcquHD9qPX1LdQFGVZ0XH3GHdwTJufnEtFz6+nC93FvKjM4ew4t5zeOTS8V7LKJg2uB/bjlS4nEHjCNgOSXLvhjk8OYajFfUBK1FR39TC2+vzOH9sMv1jvJf+2DcylLNG9OfjLfnYbO65dXYXVlHb2GICtq0QEUaknDwZSn1TC2+sPczs0cltxq2yMuKoa2pxOzFg9YEyGpptzBza9sRB3QFP0zIvE5E8YCbwiYgstq9PE5FFAEqpZuCnwGJgJ/CWUmq7Z2b7hxMlFdwTfEeJBX+4dTbkHuOqp1byvadWsulwOb84bwTf3Teb+y8Y7VWxAu3Hb7Gp48HDzth6pIIxqa4HbB2MCHCmzmfbCiivbeK6bPdH1nbGgompFFTWs+6Qe5PEd8cBV/5iZIquqeOIjSzamk9ZTSM3zsxss32WPXDr7gCsr3YVERFqYWY7M8V1BzzN0nlPKZWulApXSiUrpeba1x9VSl3o1G6RUmqEUmqoUuqPnhrtL3IKq7AIDEly7xHuhOD71q1TUdvEzS+u5VBZDQ9dNIYV957Dz2YPp2+U67Vy3GHKoH5YBJf8+C02xY6j7gVsHThcaIFy6yxcnUtmQhSn+aAnd96YZCJDrW67dTbmlpMYHUZ6v+5XodHXjEqJoaKuicLKBgBeWnmIIUl9OH1Y27/foIQo+kaGuhW4VUrx5a4iTh+a2K1jKGakbQfkFFaRmdjH7R+4b1QoA+IifS74j3+5h4q6Jl64KZubzxjstdKz7REdHsLYtL6sdiEff19xNXVNLV3KKMnoF0VEqCUggduvc4pZc7DMq8FaZ6LCQpg9uj+Ltua75UPemHuMrAwz4KotTgRuK9l8uJzNh8u5Ycagdq+ViJ4RzNUnVdAJCLlltZztYqXZYMUIfgfkFFYzoovzq45OjfFp4PZASQ0vrzzIVVMy/FoXPXtwPJsOl3daodAxwrazGvhtYbEIw/pH+z0Xf9nuIn708jpGpcRwrRdy79tjwcQ0SmsaWbnftYynYzWN7C+pYfKgOJ/Z1J05nppZWMXLKw8RFWbl8ikdB9uz0vuSU1jl8twLX+3S1WLPNYLfM9ER+RqXiqa1xZjUWPaX1FDf5PkMO23xp0U7CbNa+MXcET45fntMy4ynodl2XNDbY+uRCiJDrQx1M2DrYET/GL8K/hc7C7nt5fUMS4rm9R/NILaTyqiecNaIJGLCQ3h9jWsTyzhSCCdlGP99W8RFhZEcG87KfaV8tOUol08e0OnvlzUwDpuCbUdc65R9uauIkckx3XLSE2eM4LfD3qJqbMr1kgqtGZ0aS4tN+US0vttXwuc7CvnJOcO8HpjtjGmZWnTWHOzYrbPNXhLZ3YCtg+HJMRRWNlBR5/tMncXbC7jj1fWMTIlh4Y+m08/HoygjQq3cMmswi7YW8N2+kk7bb8w9hkVgYoYZcNUeI1Ni+Wp3MY3NNm5oJ1jrzIT0OAA2He48eF5Z38Tag2UuTxwUzBjBb4c99rlVO5vWsD0cwcr1bmZjdEaLTfHIxzsZEBfJLQGYxDohOpzh/aM7rKvTYlNsP+paSeT28Ffg9tOt+dz52gbGpvXl1VunExflnyHzd5w1lIHxUTz0wXaaOvHlbzxczqiUWJ/HaLozDrfOjCHxLmXVJUaHk94vks2HOx8E9+2eEpptqtu7c8AIfrvsLqgm1CpdHmSRER/F8P7RLNle6FW73lmfx478Su69YFTAsgWmDY5n/cFjtLSTS77fHrDtSoaOgxF+mP3qo81H+enrG5mYEccrt2S7NROYp0SEWvntgjHsLarmhW8PtNvOZlNsyi03+fed4BB8V3r3DrIy4lwacfvlriJiI0KCftJ4VzCC3w57CqsYkhjt8oTVbTFvXAqrD5RSVuNewbH2qG5o5v+W7GbSwDgWTPCsVIInTB8cT1VDc7tZSF0dYevMgLhIosKsPvPjv7/xCHe/sZEpA/vx0s3Znc5m5gtmj05mzuj+PLZ0DwXtVG/cW1xNVUOzyb/vhPkTUnniuknMG+t6zaisjDiOlNdRXNXQbhubTfHV7mLOHJHkk5r+/qb7fwMfkVNUxfAu+u8dzB2bgk3B0h3e6eU//fU+iqsaePCiMQFNz5uWqSembs+ts/VIBRGhFoa6OX7BGYtFGN4/+rhrzZv8Z30eP39rE9mD43nx5mlEhwfOVfLQRWNpsin+uGhnm9s32F2CpoffMeEhVi6akOZWKu1EFwZgbTtaQUl1Q49w54AR/DbZfLicw2V1xx8Tu8rYtFjS+0Xymb0evSccKa/jmW/2c/HEtID39tLiIknvF9mu4G+zj7D1tEc03AezX72zPo9f/Wczpw9N5IWbsgPuFx+YEMVPzh7KR5uPthnA3ZhbTlxUKIO7cf2WYGVcWl+sFulwANaXu4oQ0ZlVPQEj+K04XFbLLS+tJSM+kmuzPcvFFhHmjU1hxZ4Sj+vC/O9nuwC494JRHh3HW2QPjmftwbJTSv16I2DrYERyNMVVDW7X4G+PXQWV3P/eVmYMTuDfN04lMiw4RkzecdZQMuIj2wzgbsg9xqSMODPgygdEhlkZmRzToR//q11FZGXEkRDt/tSgwYgRfCcqapv44YtraWy28cJN07zyI88dl0Jji42vdne9tv/G3GN8sOkoP5o1JGjygLMz4ymtaWRf8ckzex0oqaa20bOArYPhXgzc1jW28LOFG4mNCOXxaycF1fD4iFArDy8Ye0oAt6KuiT1F1QF/ouvJTMyIY/Ph8jaL2RVXNbA5r4JzR/YMdw70UMF/eeVB9he7JxKNzTbueHU9h0prePoHUxnWxRG2rZk8sB+J0eEs3tY1t45Sij98vIOkmHDuOHuoV2zyBtmDtR9/bat8/OMB2y6MsG2NN2e/euSTHewpquavV03s0kTuvmb26GRmjzo5gLvF7mowM1z5jqyMvlTWN3Ow9NQpSZft1qNre0L+vYMeJ/jltY387fMcLvrHCt5ed9il2YWUUtz/7lZW7i/l0csneLX8qdUinD82ma92F3Vp1O3HW/LZkFvOL88fEdDgYmsGJ/YhMTr8FD/+tiOVRIRaGNbFEbbOpPWNIDo8xONc/M+25fPa6lxuO3MIZwaxL/a3C04O4G44VI6YAVc+Jcs+erktP/6y3cX0jwlnrB9Ll/iaHif4cVFhLLp7FhPS+/Kr/2zhrjc2tTnfpTP/+HIv72zI4+7Zw7mikxocXWHe2BRqG1tYsafzUZXO1De18OinuxiTGsv3pmR0voMfERGyB586sfnWIxWM9kLA1nEOXVOn6y6do+V13PvOVsYP6Msvzx/psU2+ZGBCFD8+60QAd+PhY4zoHxOQlNHewrD+0USFWU8ZgNXUYuObnGLOGdm/R8VPepzgA6T2jeS1W2fwq7kjWbQ1nwsfW97uiNf3Nx7hr5/ncPmkAdwzZ7hP7JkxJIHYiBC3s3VeW53LkfI6fjN/dJdLFPiS7Mx4jpTXkXesFtA5yzu8FLB1MCK566mZLTbFPW9uoqnFxuPXTiIsJPj/3H989okA7kYz4MrnWO1zBG9sFbhdd/AYVQ3NPcqdAz1U8EH/kHeeM4y375iJCFz19Eqe+HLPSaNDV+8v5b/+s4UZQ+J59IoJPruTh4VYmDM6maU7CzsdRu+gtrGZfy3by2lDEzhtWKJP7PKUExOb617+gdIaqhuavRKwdTAiOYaS6sYuDV7751d7WXOgjD9cMq7bpDVGhFr57UU6gFtR12QCtn4ga2AcO49WnlQB9qvdRYRahTOGB+f/XlfpsYLvYPLAfnxy1yzmj0/lz0tyuP7fq8ivqGNfcTW3vbKejPhInv7+VJ/3/uaOS6G8tqnDGjTOvLLyECXVjfzifP9Ww3SHkSkxxESEHP9O27wwwrY1w7sYuF13sIy/L83h0qw0Lp88wGv2+IM5Y3QAF8yAK3+QlR5HY4uNXU4z1H25q4jpgxOCKm7mDXq84APERoTy2DVZ/PnKiWzJq+CCx5Zzw3NrCLEIL9yU7bMZopw5c3gSkaFWPnMhW6e6oZmnvt7HWSOSmDIo3ue2dRWrRZiWGX9c8LfmVRAeYmF4f88Dtg66UkStoraJu9/YRHq/KP5w6bhu6YN99IoJPHr5eIZ58Voa2sYx4taRj3+4rJa9RdU9zp0DvUTwQQcAvzclnU/umkVGvyhKaxp49sapDEzwzgTfnREZZuWsEUks3l7Q6QTWL313kGO1Tfz8vODt3TuYlhnPvuIaSqobvBqwdZASG0FMeIjLgVulFL9+byuFlfU8dk1Wtw14JsWEc032wG55s+pupPaNICkm/HiJhS/tk52cMzJ4M7q6Ss96XnGBwYl9eO8np1FV3+zzuuetmTcuhc+2F7DxcDlTBrXtm62sb+KZb/YzZ3T/45MtBzOOfPw1B8rYfrSSyyZ5130iIgxPdn32qzfXHuaTrfncO2+UyV83uISI6MqZ9tTML3cVkZkQxRAvpBYHG72mh+9MiNXid7EHPYAj1Cos6SBb5/kVB6ioa+KeOcHfuwftr48ItfDWusNUNzR71X/vYESynv2qszEVe4uq+N1HOzh9WAK3nznE63YYei5ZGXHsL66hsLKelftLe6Q7B3qp4AeKvpGhnDY0kc+2F7QpXuW1jTy3/ADzxqZ4NdPFl4SFWJiU0Y9l9tIRvrB7eHIMx2qbKKluP1OnvqmFny7cSGSYlb9eleWTCcgNPZeJ9hmwnvp6H43Nth5THbM1RvD9zLxxKRwqrWVXwakuimeX76e6sZl7zvPNeABf4XDrhIVYPC4p3RauBG7/e9FOdhVU8ZcrJ5Ic699pHw3dnwn20cyvrc4lKsx6/G+6p2EE38+cNyYZEU7J1imraeSFbw8yf3wqo1K611Buxz/H6NRYjyaMaY/Oaup8vqOQl1Ye4ubTB/fYR3GDb4mNCGVoUh8am22cMSyR8JDgKa7nTYzg+5nE6HCmZcazuJUf/+mv91Hf1OKz0b6+ZNLAOMKsFiZ6oWBaW/SPCSc2IoScolMzdfIr6vjVfzYzNi2Wey8I7tIJhuDGkZ7ZkzsNRvADwNyxKewqqOJAia7QV1zVwEsrD3JJ1gCvVen0J1FhIbxx+wzunu2bm5WIMCI55hSXTotNcc8bm2hstvGPayf12F6ZwT+cNjSR8BBLj/XfgxH8gDB3bDLA8V7+v5bto6lFcZePBNMfTB7Yz6eTRDhmv3IOdv/zq72sPlDG7y8Z1yNT6Az+5fJJA1h5/+weHQMygh8A0vtFMX5AXz7bVkBhZT2vrj7E5ZMGdJt6L4FgRHI0FXVNxyecdpROuCQrjSu6WekEQ3BisQjxAUjX9idG8APEvHEpbDpczkMfbMNm6969e38w0mn2K+fSCY9009IJBkMgMIIfIOaOTQFg8fZCrpyaQUa8f0o8dFeci6jd9+4WCivrefzaSd22dILBEAh6XWmFYGFY/2iG9Y8mt7SWn547LNDmBD2J0WH0iwrlmW/2U1BZz/0XjOoWpScMhmDCCH4A+d3FYymtaQyaicmDGV1TJ4Y1B8qYNTyRH80ypRMMBncxgh9ATg/SiU2ClckD+3GwpIa/XDXRlE4wGLqAuDLJd6CYOnWqWrduXaDNMAQJNpuiodlGZJjJtzcY2kNE1iulpra1zaOgrYhcKSLbRcQmIm2ewN7uoIhsFZFNImIU3NAlLBYxYm8weICnLp1twOXA0y60PUcpVeLh+QwGg8HQRTwSfKXUTsDkQRsMBkM3wF95+ApYIiLrReS2jhqKyG0isk5E1hUXF/vJPIPBYOj5dNrDF5GlQEobmx5QSn3g4nlOV0odFZH+wOcisksp9U1bDZVSzwDPgA7aunh8g8FgMHRCp4KvlJrj6UmUUkft70Ui8h6QDbQp+AaDwWDwDT536YhIHxGJcSwD56ODvQaDwWDwI56mZV4mInnATOATEVlsX58mIovszZKBFSKyGVgDfKKU+syT8xoMBoPBfTzN0nkPeK+N9UeBC+3L+4GJnpzHYDAYDJ5jqmUaDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAbDAZDL8EIvsFgMPQSjOAHO0oF2gKDwdBDCAm0AQagvhKOHYCy/fbXAftrP9SXw9n3wWl3gUigLTUYDN0YI/iBZOWTsPwvUFty8vo+/SF+CAw5G2qK4fOH4OhGuOSfENYnIKYaDIbujxH8QLLpNYjoC6f9VAt8/BDolwnhMSfaKAXf/h2++D0U74arX4WEoYGyuOeTsxiW/1XfXBOHBdoag8GrGB9+oGiqg6KdMPYyOOPnMOYSSBl/stiDduOc8XP4/jtQlQ/PngM5SwJjc0/G1gJf/QkWXgWHV8GKvwXaIoPB6xjBDxSFO0C1QOpE19oPPRduWwZxA7Uoff2/YLP51MReQ20ZvHYlfP0/MPE6yPo+bH0LqosDbZnB4FWM4AeK/I36PS3L9X36ZcLNS2DCVfDVH+HN70N9hS+s6z0c3QhPnwUHl8NFf4NLn4Qz7oGWRlj3XKCtMxi8ihH8QHF0E0TGQ98M9/YLi4LLnoZ5/wM5n8Gz50Lhdp+Y2OPZ8Ao8NxeUDW7+DKberF1oicNh+Pmw9t/QVB9oKw0Gr2EEP1Dkb9bunK6kWorAjDvgxo90D/+pM+C9O3QqZyBoboD374TvntC+cE+O4w83VVM9fHgXfPhTGDQTbv8GBkw5uc2Mn+gMqW3v+N4eg8FPmCydQNDcoAO2p/3Us+Nkng4/WQ3f/g3WPAtb34as6+HMX0Gcm08OXUUp+Ohu2Py6/rzrE7jsX9r95CrNDfDd4/DNX/TnhGGQNAISR+redtJIiB8KoRFdt7OxFqoLoOIIfP6gduXM+gWc8wBYrKe2H3I29B8Dq56ErOvMGAhDj8AIfiAo3A62JkjN8vxYfRLg/Edg5k91OuH6F7T4TrlJC1pMiufn6IhvH9PnO/vX+iaz6L/gX6fDvEdh0vc7F8p9X8GiX0LpXhi9AOIG6fTTvHWw7V3APtJYLHpbXAaEREJIOIREnPweGgmWUKgrg+pCqC6CqgL93lh14pzhsXDNQhg1v327RGDGj+HDn8GBb2DIWR5fKoMh0IgK4qH7U6dOVevWrQu0Gd5n3Qvw8T1w1yaIH+zdY5cfhuV/ho2vgiUEpt2q0zr7JHr3PAC7FsEb18G4y+GK57RIlufC+z/RQdCR82HBYxCddOq+lfmw+New/V09/uDC/4Nhc05u01gLZfv0DaBkD5Tshsqj0FyvnwpavzfVAQrCYiAmGaKTIbo/RKfY35P1+pQJ+nNnNNXD38ZC+jS47g2vXDKDwdeIyHql1NQ2txnBDwAf3Q3b34d7D/rOVVC2X6dubnkTQqNg7h9h8o3eO1/BNnjufO1u+eEi3bt2YLNpV8gXv9fjCi5+/ERvuqUZ1jyjc95bGvVTyOl3e+aucaCUjiFYvfjg+tWf9HX82XrvDXhrboT3boPQPnDuAxCb5p3jGgx0LPgeBW1F5P9EZJeIbBGR90Qkrp1280Rkt4jsFZH7PDlnj+Dopq4HbF0lfghc9pT28Q+Yom8yC6+GqkLPj11dBK9fo0cJX7PwZLEHsFh0fOK2ZRCbqp8CPrgT9n0Jz5wFi++HgdPhzlVw9r3eEXvQ19ObYg8w9RawhsKqf3nneI6Yx/b3dK7/P6boG0pTnXvHKdoJix+Agyu8Y5ehV+Bpls7nwDil1AQgB7i/dQMRsQL/BC4AxgDXisgYD8/bfWluhKIdrg+48pSkEfCD93Ua54Gv4ckZsOODrh+vqR7euB5qSuDa17Wgt0fyGLj1S92L37QQXrkM6o7BVa/A9f/RN6VgJyYZxn1Pl8GoO+b58Vb8FTYvhLPug5+uheHn6TEV/5gKW//TcXVUWwvs/BheWqB/x5VPwOvXaXeXweACHgm+UmqJUqrZ/nEVkN5Gs2xgr1Jqv1KqEXgDuMST83ZrindqV4Y7A648xWLRaZy3f6NH6r51g07jdHfQlqN3mrdGPz248h1CwmD2Q3DzYjjv93DnGhhzcffKepn5E2iqhQ0ve3ac7e9rN9e47+kKqP0y4aqX4aZFEBUP79yi3WR560/er7ZMB8cfy4I3r4fS/TD7t3Db1/rp4/VroK7cM9t6A7sWQcHWQFsRULyZh38z8Gkb6wcAh50+59nX9U6ObtLv3sjQcZekkXDrUjjrXtjyls6mOfCN6/uv+BtseUOnMo691L1zZ2RrX314tHv7BQMp4yFzFqx+RscgusKR9fDe7ZAxXRdmc77hZZ6u3V8XPwHHDsK/z4V3b4cDy3WW0F/H6IqpcQP109Hdm2HW/9M33Ktf0fu8c4tnYyB6Oge/1a7F1691333Wg+hU8EVkqYhsa+N1iVObB4Bm4LW2DtHGunafW0XkNhFZJyLriot7YC2T/M06LbCfl7NzXMUaCuf8Gm5ZAtYw7R747Nc6a6axtn2Xws6P7b3TK3Sef29jxk+gMg92dsEdVn5YC010f7j6tbZjFhYrTP4B3LVBZ1Vtfxdeugi2vA0TroQ7voUffqKfjpzjFINOgwv/DHuXwtLfdv379WTqK+H9OyAqASoOey8e0w3xOEtHRG4E7gBmK6Vq29g+E3hYKTXX/vl+AKXUf3d27B6ZpfPsuTpr5qaPA20JNNbonuPaf59YZwnRmTXhsRARq9/DY/WTQFsZOb0Fmw2emKJF49alru/XUAXPz9Ppqrd8Dv1HubbfsYOQu0qXeIiK77z9J7+Etc/qshsTr3Hdvt7AB3fqGNLN9tLXB1fAXRvbThfuAXSUpeNRSoOIzAPuBc5qS+ztrAWGi8hg4AhwDXCdJ+fttrQ06XTG7B8F2hJNWB+Y/xeYcDUUbNHiVF+p3xsqTyxX5sGAyXDFv3un2IOOg0z/MXz6Kzi8FjKmdb6PrQXeuVVn1Fz/tutiD9q/785o5Xn/DcW7dMmIhOGQPqXzfXoDuz7RY1Jm/UK7Fc//gw54L/uTLpbXy/A0h+0JIBz4XLRPcpVS6g4RSQP+rZS6UCnVLCI/BRYDVuB5pVTvrPZVvBtaGgLjv++IjGz9MnRM1nXw5SOw6p+Q8WLn7Zc8qAvczf8LDJvtW9usoXDlS/Ds2dpX7UiJ7c1UF+sbYMoEnRUFulTH1Jv1U232bdB/dGBtbIvGWj3IMG2S1w/taZbOMKVUhlIqy/66w77+qFLqQqd2i5RSI5RSQ5VSf/TU6G5L/ib97s8MHYP3CI+GKTfAjg+1X74j1j2vbwzTf6xHO/uDPglw7Rv6qezN63t3pU9HRllDFVz+jM4Wc3DWfXo09pIHA2dfexxYDv86DV69QrtcvYyppeNP8jdDWLQuBGbonmTfrucifv0aXaeouUG76lrs780NOu224jAMn6tHOPuT5LFw+dN6roSP7tbps90pBRa0K6ymRBe7O14PqRCSx8GIua59n02vwe5P4Pw/ntqL75MAZ/5SF9Hb+4Xvn75cob5Cx9PWv6gTOq58ySfzVxvB9ydHN+nHS4upSt1ticvQWTR7FuuBWNZwXbgtIlZnPTle0ZfAWf/VdiVOXzN6gU6d/eqP+gZw+l3+t8Edast0+mn5IT0SvLZEz1HQFsPP13WXOopvHDsEn94Hg87Q2VVtMf127dZZ8htdGTUQv5ODnMXw0T36Bjfzp/q3C4vyyamM4PuLlmY96GPqDwNticFTZj+oX8HMmb+Cwm2w9GEtku4EjP3Nxldg18fazrRJrYrd2ZejEvTAt6/+BP+crr/faXed7KoB/XTw/o/18mX/ar9zFRKuBwK+faM+/5SbPPsONpu9iF+9vnlE9O18n5pS+Ow+XWIjaTRc/arPg+1G8P1F6R5orvNfSQVD70YE5v9Vl5/+/CG4/q1AW9Q2SumUyfRsncnUETPvhDGXapH88g+6MOD8v8LgWSfarPwnHPoWLnlSD1TriDGXQMYM+PKPenxJeEzH7Uv26Bto6T79v9xUf+K9peHkttHJ2pWUNFrfbPuP0WnNEX31d97+Hiz6FdSX65jCrF+cevPyAUbw/UUgR9gaeid9Eu2+6oe08A89J9AWncrRjTqd9KK/u9a+7wA9ujhniZ5H4aWLYMI1ek6ImiJ9Ixh1kc6o6gwRHWP592xY8ff2n9qaG/T25X/WacmZs/RYmtAIPTeD83tolO7lF+/WNbM2vKTLcjiIHaCfVgq26KeZSz7Ubjc/YQTfX+Rv0n8MicMDbYmhN5Ht5Ku+/ZvA+qrbYtNCPYHN2Mvc22/E+ZC5Sovwt49DzqcQ2U/3oBc85nqgOn0qjL9SF6Kb+kPo26oc2KHvtH+9ZDeMvVxP7BOT7LqdNhtU5ELRLn0DKN6lpyI97w86vuDt6q6dYATfX+Rv1jVZgu0fztCzCY2AOQ/Df27WmSuTbwi0RSdobtDTco66CCLj3N8/LEoX5ptwNXz8/7Qr59rX3Z/sZ/ZDOtX2i9/rFE7QAfnPf6t76H0H6uquw89z30aL5cQgupHz3N/fyxjB9we2Fsjfoqf8Mxj8zdjLdf2YLx/Ry8FSwG73p9qHnXWtZ8dJGqlLlVQXudf7dhA3UMcHVvwVpt8Bxw7oLJ/aUnvWzK99kiIZCEx+oD8o3QtNNSZgawgMIjD3TzqX/bvHA23NCTYthJhUGOKF2IJI18TewRk/hz5J8NLF+mkoNg1u+0r7+HuI2IMRfP+Qv1m/mxG2hkCRka17998+DhVHAm2Nzrffu1QXegsGN2dErA78WqzaT/+jL3tkB80Ivj84ukkHphJHBtoSQ29mzm9BtWjXTqDZ+pa2ZWIQ1VGceI2eZ3rGj4PjJuQDep7gN9bAF3+APZ8H2pIT5G/Ww8L9HJE3GE6iX6YWs80LT6QJB4LjuffT9BScwUR3K0PhJj1P8K3hejanbx8LtCUam00LvnHnGIKBWb/QeeBLftPx/Lm+JH+TTlGc6GGw1uA2PVDwQyD7Vji4XNeeDzRl+6Gxygy4MgQHEX3h7Pv1/8fuRYGxYdNC3TEbd3lgzt+L6XmCDzrXODQK1jwdaEtOlETugQEgQzdlyk2QOEKPwG1p8u+5j+fez9cDpQx+pWcKfmQ/PRhjy1u6QFEgyd+kqycG40QLht6JNVSP9Czdq+v2+5Ocz/Sgpqzr/XteA9BTBR90+dPmej1SLpAc3aRrZVhDA2uHweDMiLkw+CxY9t9agP3Fptd17n0w1vXpBfRcwe8/Wv9Br/23Lk0cCJTSI2yN/94QbDgKh9VXwMuX6AqQvqa6CPYs0U/fPTTtMdjpuYIPOgWt8gjs+igw5z92ABoqTIaOIThJGa9rsB87BE+fqV2gvmSLPffelUqWBp/QswV/+Pk693h1gIK3x0sim4CtIUgZNR/uWKHF/90fwfs/8clcqjr3/jUYMEXXvjEEhJ4t+Barnpk+d2VgBprkbwZLqJ78wGAIVuIy4MaP4cz/0imTT5+lZ2fzJvmbde696d0HlJ4t+KCzAUL7wJpn/H/u/E2QPEZPp2YwBDPWEDj3AbjxQ2iogmdnw5pnvTc4a/PrOltt3BXeOZ6hS/R8wY+M072KrW9DdbH/zquUfqow7hxDd2LwmfDjb2HIWXpGqTe/rycZ94TmRu2/N7n3AafnCz5ot05LI6x/0X/nLN6ta32bDB1Dd6NPIlz7pi6pnLNYu3iKdnb9eHsWQ12Zyb0PAnqH4CeNgKGz7SmafhpZuO457b8fvcA/5zMYvInFoicFuWWx7iw9dz7s/cL94xRsgyUPQnSKd+reGzyidwg+6Jlsqgtgxwe+P1d9hQ5+jbsCovv7/nwGg68YMAV+9IWeFeq1K10fmasUrHtBTxDeVAdXvmiqxQYBvUfwh82B+KGw+infn2vja9BYDTPu8P25DAZf0zcdbv4Mhs2Gj38Oix/Q03a2R0MVvHMrfHwPDJyp0z4HzfSbuYb26T2Cb7Hocgt5ayFvve/OY2vRRdsypkPaJN+dx2DwJ+ExcM3rOh628gl464a28/ULtmqf//Z34dwH4fvvQnSS/+01tEnvEXzQ9bfDYnxbRXPPEjh2ULuQDIaehDUELvw/uOB/dWnlFy6Eyny9TSnt7nl2NjTV2vP6f6k7WoagoXf9GhGxMOl62PYuVBX45hyrn4LYASZYa+i5TL8drn0DSvZoH/2hlfDOLdrdk3mGduFknh5oKw1t0LsEH/Qjqa1ZB5S8TdFO2L8Mpt1iqmMaejYj5mq/vlLwwjzY/h7Mfgiu/49O6zQEJb1P8BOG6ho7656DhmrvHnv103qy8sk3efe4BkMwkjoBfvSlnnDopk/09InGhRPU9M5f58xfQU0xfPO/3jtmbRlsfgPGXwl9Erx3XIMhmIlNhYv/AYNOC7QlBhfonYKfMQ2yvg8r/6lHxHqDDS9Dc50uyWwwGAxBiEeCLyL/JyK7RGSLiLwnInHttDsoIltFZJOIrPPknF5jzsMQ1gcW/crzAlEtzXoUb+YsPbuVwWAwBCGe9vA/B8YppSYAOcD9HbQ9RymVpZSa6uE5vUN0ks4TPvA17Hjfs2Pt/gQqDptUTIPBENR4JPhKqSVKKcf8gauAdM9N8iNTb4aUCXrkoCcB3NVP66HnIy/wnm0Gg8HgZbzpw78Z+LSdbQpYIiLrReQ2L57TMyxWmP8XPQ3iN//XtWPkb4FD3+p0TzNPp8FgCGI6FXwRWSoi29p4XeLU5gGgGXitncOcrpSaDFwA3CkiZ3ZwvttEZJ2IrCsu9kP9+oxsXbZ15T+hOMf9/Vc/DaFRMOkH3rfNYDAYvEingq+UmqOUGtfG6wMAEbkRuAi4Xqm2o59KqaP29yLgPSC7g/M9o5SaqpSampTkpxocc36nRftTNwO4NSV6YpWJ1+qJVgwGgyGI8TRLZx5wL3CxUqq2nTZ9RCTGsQycD2zz5LxeJzoJzv2NHiXrTvnk9S9AS4MJ1hoMhm6Bpz78J4AY4HN7yuVTACKSJiKL7G2SgRUishlYA3yilPrMw/N6n6k3Q8p4WPzrtqsAtqalCdY+pydWSRrhe/sMBoPBQzyakUApNayd9UeBC+3L+4Hgn9jVGgIX/gWeP18HcOc83H7b+gpY/QxU5cOCx/1mosFgMHiCmYLGmYHTdQD3uyf0e+LwE9uqi2DXJ7DrY9j/NdiaIH2anljFYDAYugFG8Fsz53ew82M9AnfB3/Xyro8hdxWgoF+mLg87eoEWfFMsymAwdBOM4LcmOgnOfQA+/S94zO6JSh4HZ92rRT55LIgE1kaDwWDoAkbw22LqLbpUQnQyjJoP8UMCbZHBYDB4jBH8trCGwPmPBNoKg8Fg8CrGAW0wGAy9BCP4BoPB0Eswgm8wGAy9BCP4BoPB0Eswgm8wGAy9BCP4BoPB0Eswgm8wGAy9BCP4BoPB0EuQduYsCQpEpBg4FGg7OiARKAm0ES7QXeyE7mOrsdP7dBdbg93OQUqpNmePCmrBD3ZEZJ1Samqg7eiM7mIndB9bjZ3ep7vY2l3sbAvj0jEYDIZeghF8g8Fg6CUYwfeMZwJtgIt0Fzuh+9hq7PQ+3cXW7mLnKRgfvsFgMPQSTA/fYDAYeglG8A0Gg6GXYAS/FSLyvIgUicg2p3UTRWSliGwVkY9EJNa+/noR2eT0solIln3bMhHZ7bStfwDtDBWRl+zrd4rI/U77TLGv3ysij4t4d/5GL9oZTNczTEResK/fLCJnO+3j0+vpZVt9fU0zROQr+2+5XUTutq+PF5HPRWSP/b2f0z7326/dbhGZ67TeZ9fVy3b69Jp6jFLKvJxewJnAZGCb07q1wFn25ZuBP7Sx33hgv9PnZcDUYLATuA54w74cBRwEMu2f1wAzAQE+BS4IUjuD6XreCbxgX+4PrAcs/rieXrbV19c0FZhsX44BcoAxwP8C99nX3wf8j315DLAZCAcGA/sAq6+vq5ft9Ok19fRlevitUEp9A5S1Wj0S+Ma+/DlwRRu7Xgu87kPTTsJNOxXQR0RCgEigEagUkVQgVim1Uum/1peBS4PNTm/a4yU7xwBf2PcrAsqBqf64nt6y1ds2tYVSKl8ptcG+XAXsBAYAlwAv2Zu9xIlrdAn6ht+glDoA7AWyfX1dvWWnt+zxJUbwXWMbcLF9+Uogo402V3Oq4L9gf6x70BeP9m3Qnp3/AWqAfCAX+LNSqgz9R53ntH+efV2w2ekgWK7nZuASEQkRkcHAFPu2QF3PrtjqwC/XVEQygUnAaiBZKZUPWmzRTx6gr9Vhp90c189v19VDOx34++/UZYzgu8bNwJ0ish79yNfovFFEpgO1SqltTquvV0qNB2bZXz8IoJ3ZQAuQhn4E/YWIDEE/HrfGH3m67toJwXU9n0f/k68D/g58BzQTuOsJ7tsKfrqmIhINvAPco5Tq6Imtvevnl+vqBTshMH+nLhMSaAO6A0qpXcD5ACIyApjfqsk1tOrdK6WO2N+rRGQhWsxeDpCd1wGfKaWagCIR+Rb9WL8cSHc6RDpw1Jc2dtHO/cF0PZVSzcDPHe1E5DtgD3CMAFzPLtrql79REQlFi+hrSql37asLRSRVKZVvd9cU2dfncfLTh+P65eHj6+olOwPyf+8OpofvAo5Iu4hYgN8ATzlts6Afod9wWhciIon25VDgIvQjd6DszAXOFU0fYAawy/6YWiUiM+yPnjcAHwSbncF2PUUkym4fInIe0KyU2hGo69kVW/1xTe3X4Dlgp1Lqr06bPgRutC/fyIlr9CFwjYiE291Pw4E1vr6u3rIzUH+nbhHoqHGwvdA99XygCX0nvwW4Gx25zwEexT5C2d7+bGBVq2P0QWdDbAG2A49hj+IHwk4gGnjbbssO4FdOx5mK/qPcBzzh/N2Cxc4gvJ6ZwG50cG8puhytX66nt2z10zU9A+3q2AJssr8uBBLQgeQ99vd4p30esF+73Thl4vjyunrLTn9cU09fprSCwWAw9BKMS8dgMBh6CUbwDQaDoZdgBN9gMBh6CUbwDQaDoZdgBN9gMBh6CUbwDQY79vz/FSJygdO6q0Tks0DaZTB4C5OWaTA4ISLj0GMBJgFWdE72PKXUvi4cy6qUavGuhQZD1zGCbzC0QkT+F13ErY/9fRC6/HUI8LBS6gN7ka1X7G0AfqqU+k50vfnfogdGZSmlxvjXeoOhfYzgGwytsJci2IAuQPYxsF0p9aqIxKHrsk9Cj8y0KaXqRWQ48LpSaqpd8D8BxildOtdgCBpM8TSDoRVKqRoReROoBq4CFojIL+2bI4CB6GJZT4ie4awFGOF0iDVG7A3BiBF8g6FtbPaXAFcopXY7bxSRh4FCYCI6+aHeaXONn2w0GNzCZOkYDB2zGPiZYyILEZlkX98XyFdK2dA1z60Bss9gcBkj+AZDx/wBCAW2iJ40/A/29U8CN4rIKrQ7x/TqDUGPCdoaDAZDL8H08A0Gg6GXYATfYDAYeglG8A0Gg6GXYATfYDAYeglG8A0Gg6GXYATfYDAYeglG8A0Gg6GX8P8BbPmGsNND930AAAAASUVORK5CYII="
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "source": [
    "exploration('qikB_d18O_raw.csv', 'qikB_SODA_1x1_raw.csv')"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "pearson correlation:\n",
      "[[ 1.         -0.34089223]\n",
      " [-0.34089223  1.        ]]\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "source": [
    "algae_site_info = pd.read_csv(\"algae_site_info.csv\")\n",
    "algae_site_info"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "         site    lat    long  depth_meter\n",
       "0  arctic_bay  73.25  274.75            5\n",
       "1    amchitka  51.42  179.23            5\n",
       "2        attu  52.78  173.17            5\n",
       "3    kingitok  55.40  300.15           15\n",
       "4       maine  44.13  291.75           10\n",
       "5        qikA  67.56  295.98            5\n",
       "6        qikB  67.56  295.98            5"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>site</th>\n",
       "      <th>lat</th>\n",
       "      <th>long</th>\n",
       "      <th>depth_meter</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arctic_bay</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>amchitka</td>\n",
       "      <td>51.42</td>\n",
       "      <td>179.23</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>attu</td>\n",
       "      <td>52.78</td>\n",
       "      <td>173.17</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kingitok</td>\n",
       "      <td>55.40</td>\n",
       "      <td>300.15</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>maine</td>\n",
       "      <td>44.13</td>\n",
       "      <td>291.75</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>qikA</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>qikB</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 155
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "By the way, here is where our sample locations are in the world. You can see that they're a bit crowded around the Northwest of the map below. "
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "source": [
    "import shapely\n",
    "from shapely.geometry import Point\n",
    "import geopandas as gpd\n",
    "from geopandas import GeoDataFrame\n",
    "\n",
    "algae_site_info['long_formatted'] = algae_site_info['long'] - 360\n",
    "algae_site_info"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "         site    lat    long  depth_meter  long_formatted\n",
       "0  arctic_bay  73.25  274.75            5          -85.25\n",
       "1    amchitka  51.42  179.23            5         -180.77\n",
       "2        attu  52.78  173.17            5         -186.83\n",
       "3    kingitok  55.40  300.15           15          -59.85\n",
       "4       maine  44.13  291.75           10          -68.25\n",
       "5        qikA  67.56  295.98            5          -64.02\n",
       "6        qikB  67.56  295.98            5          -64.02"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>site</th>\n",
       "      <th>lat</th>\n",
       "      <th>long</th>\n",
       "      <th>depth_meter</th>\n",
       "      <th>long_formatted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arctic_bay</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>-85.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>amchitka</td>\n",
       "      <td>51.42</td>\n",
       "      <td>179.23</td>\n",
       "      <td>5</td>\n",
       "      <td>-180.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>attu</td>\n",
       "      <td>52.78</td>\n",
       "      <td>173.17</td>\n",
       "      <td>5</td>\n",
       "      <td>-186.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kingitok</td>\n",
       "      <td>55.40</td>\n",
       "      <td>300.15</td>\n",
       "      <td>15</td>\n",
       "      <td>-59.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>maine</td>\n",
       "      <td>44.13</td>\n",
       "      <td>291.75</td>\n",
       "      <td>10</td>\n",
       "      <td>-68.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>qikA</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>-64.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>qikB</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>-64.02</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 156
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "source": [
    "geometry = [Point(xy) for xy in zip(algae_site_info['long_formatted'], \n",
    "                                    algae_site_info['lat'])]\n",
    "\n",
    "gdf = GeoDataFrame(algae_site_info, geometry=geometry)\n",
    "\n",
    "#this is a simple map that goes with geopandas\n",
    "world = gpd.read_file(gpd.datasets.get_path('naturalearth_lowres'))\n",
    "gdf.plot(ax=world.plot(figsize=(10, 6)), marker='o', color='red', markersize=15);"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Introducing, Aspects of Machine Learning"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "source": [
    "temp_and_algae['lat'] = 73.25\n",
    "temp_and_algae['long'] = 274.75\n",
    "temp_and_algae = temp_and_algae.reset_index()\n",
    "temp_and_algae.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "   index  year      temp      d18O    lat    long\n",
       "0      0  1967 -1.301917 -0.132046  73.25  274.75\n",
       "1      1  1968 -0.175512 -0.134891  73.25  274.75\n",
       "2      2  1969 -0.575391 -0.137737  73.25  274.75\n",
       "3      3  1970 -1.314653 -0.134845  73.25  274.75\n",
       "4      4  1971 -1.344812 -0.131952  73.25  274.75"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>year</th>\n",
       "      <th>temp</th>\n",
       "      <th>d18O</th>\n",
       "      <th>lat</th>\n",
       "      <th>long</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1967</td>\n",
       "      <td>-1.301917</td>\n",
       "      <td>-0.132046</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1968</td>\n",
       "      <td>-0.175512</td>\n",
       "      <td>-0.134891</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1969</td>\n",
       "      <td>-0.575391</td>\n",
       "      <td>-0.137737</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1970</td>\n",
       "      <td>-1.314653</td>\n",
       "      <td>-0.134845</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1971</td>\n",
       "      <td>-1.344812</td>\n",
       "      <td>-0.131952</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 158
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "As a proof of concept, let's build a little 'neural network' with all of our algae datasets. We want to predict `temp`, given `year`, `d18O`, `lat`, and `long`.\n",
    "\n",
    "First, let's configure all the data into one big dataframe.\n",
    "\n",
    "Then, let's scale everything between 1 and 0"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "source": [
    "paleo_df = ''\n",
    "paleo_df"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "''"
      ]
     },
     "metadata": {},
     "execution_count": 159
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "source": [
    "algae_files = [['ab_d18O_raw.csv', 73.25, 274.75, 5],\n",
    "                ['amchitka_d18O_raw.csv', 51.42, 179.23, 5], \n",
    "                ['attu_d18O_raw.csv', 52.78, 173.17, 5],\n",
    "                ['Kingitok_d18O_raw.csv', 55.4, 300.15, 15],\n",
    "                #['maine_d18O_raw.csv',44.13,291.75,10],\n",
    "                ['qikA_d18O_raw.csv', 67.56,295.98,5],\n",
    "                ['qikB_d18O_raw.csv', 67.56,295.98,5]\n",
    "               ]\n",
    "                 \n",
    "                 \n",
    "temperature_files = [\n",
    "    'ab_SODA_1x1_raw.csv', \n",
    "    'amchitka_SODA_1x1_raw.csv',\n",
    "    'attu_SODA_1x1_raw.csv',\n",
    "    'kingitok_SODA_1x1_raw.csv', \n",
    "    #'maine_SODA_1x1_raw.csv',\n",
    "    'qikA_SODA_1x1_raw.csv', \n",
    "    'qikB_SODA_1x1_raw.csv']"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "source": [
    "paleo_df = ''\n",
    "paleo_df = pd.DataFrame()\n",
    "\n",
    "count = 0\n",
    "for temperature_file in temperature_files:\n",
    "    temp_data = pd.read_csv('temp_raw/' + temperature_file)\n",
    "    print(temperature_file)\n",
    "    # locate the year value\n",
    "    temp_data['year'] = temp_data['time'].apply(lambda x: x[0:4])\n",
    "    # group data annually\n",
    "    actual_annual_temp = temp_data.groupby('year')['temp'].agg(np.mean).to_frame().reset_index()\n",
    "    # convert year to an int type\n",
    "    actual_annual_temp['year'] = actual_annual_temp['year'].apply(lambda x: int(x))\n",
    "    \n",
    "    algae_file = pd.read_csv('d18O_raw/' + algae_files[count][0])\n",
    "    \n",
    "    if algae_files[count][0] == \"amchitka_d18O_raw.csv\":\n",
    "        algae_file = algae_file[['Year', 'AM-KR-80_d18O']]\n",
    "        algae_file['d18O'] = algae_file['AM-KR-80_d18O']\n",
    "        algae_file['year'] = algae_file['Year'].apply(lambda x: int(x))\n",
    "        del algae_file['AM-KR-80_d18O']\n",
    "        del algae_file['Year']\n",
    "    \n",
    "    elif algae_files[count][0] == \"attu_d18O_raw.csv\":\n",
    "        algae_file = algae_file[['year', 'd18O']]\n",
    "    else:\n",
    "        algae_file['year'] = algae_file['year'].apply(lambda x: int(x))\n",
    "    print(algae_files[count][0])\n",
    "    \n",
    "    \n",
    "    temp_and_algae = pd.merge(actual_annual_temp, algae_file, on='year')\n",
    "    #temp_and_algae = temp_and_algae.set_index('year').head()\n",
    "    temp_and_algae['lat'] = algae_files[count][1]\n",
    "    temp_and_algae['long'] = algae_files[count][2]\n",
    "    temp_and_algae['depth'] = algae_files[count][3]\n",
    "    temp_and_algae['location'] = temperature_file.split('_')[0]\n",
    "    paleo_df = pd.concat([paleo_df, temp_and_algae], axis= 0)\n",
    "    \n",
    "    count += 1\n",
    "    \n",
    "paleo_df"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "ab_SODA_1x1_raw.csv\n",
      "ab_d18O_raw.csv\n",
      "amchitka_SODA_1x1_raw.csv\n",
      "amchitka_d18O_raw.csv\n",
      "attu_SODA_1x1_raw.csv\n",
      "attu_d18O_raw.csv\n",
      "kingitok_SODA_1x1_raw.csv\n",
      "Kingitok_d18O_raw.csv\n",
      "qikA_SODA_1x1_raw.csv\n",
      "qikA_d18O_raw.csv\n",
      "qikB_SODA_1x1_raw.csv\n",
      "qikB_d18O_raw.csv\n"
     ]
    },
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "    year      temp      d18O    lat    long  depth location\n",
       "0   1967 -1.301917 -0.132046  73.25  274.75      5       ab\n",
       "1   1968 -0.175512 -0.134891  73.25  274.75      5       ab\n",
       "2   1969 -0.575391 -0.137737  73.25  274.75      5       ab\n",
       "3   1970 -1.314653 -0.134845  73.25  274.75      5       ab\n",
       "4   1971 -1.344812 -0.131952  73.25  274.75      5       ab\n",
       "..   ...       ...       ...    ...     ...    ...      ...\n",
       "22  2004 -1.368374 -1.801891  67.56  295.98      5     qikB\n",
       "23  2005 -1.031412 -2.045915  67.56  295.98      5     qikB\n",
       "24  2006 -0.294162 -2.200304  67.56  295.98      5     qikB\n",
       "25  2007 -0.349986 -2.143647  67.56  295.98      5     qikB\n",
       "26  2008 -0.626585 -2.214013  67.56  295.98      5     qikB\n",
       "\n",
       "[233 rows x 7 columns]"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>temp</th>\n",
       "      <th>d18O</th>\n",
       "      <th>lat</th>\n",
       "      <th>long</th>\n",
       "      <th>depth</th>\n",
       "      <th>location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1967</td>\n",
       "      <td>-1.301917</td>\n",
       "      <td>-0.132046</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1968</td>\n",
       "      <td>-0.175512</td>\n",
       "      <td>-0.134891</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1969</td>\n",
       "      <td>-0.575391</td>\n",
       "      <td>-0.137737</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>-1.314653</td>\n",
       "      <td>-0.134845</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1971</td>\n",
       "      <td>-1.344812</td>\n",
       "      <td>-0.131952</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2004</td>\n",
       "      <td>-1.368374</td>\n",
       "      <td>-1.801891</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2005</td>\n",
       "      <td>-1.031412</td>\n",
       "      <td>-2.045915</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>2006</td>\n",
       "      <td>-0.294162</td>\n",
       "      <td>-2.200304</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>2007</td>\n",
       "      <td>-0.349986</td>\n",
       "      <td>-2.143647</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2008</td>\n",
       "      <td>-0.626585</td>\n",
       "      <td>-2.214013</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>233 rows × 7 columns</p>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 161
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "source": [
    "dataset = paleo_df.values\n",
    "dataset"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([[1967, -1.3019167168272867, -0.13204552, ..., 274.75, 5, 'ab'],\n",
       "       [1968, -0.1755116544922609, -0.13489132, ..., 274.75, 5, 'ab'],\n",
       "       [1969, -0.5753912617866366, -0.13773713, ..., 274.75, 5, 'ab'],\n",
       "       ...,\n",
       "       [2006, -0.294161577132486, -2.200303795, ..., 295.98, 5, 'qikB'],\n",
       "       [2007, -0.34998577832898486, -2.1436467009999998, ..., 295.98, 5,\n",
       "        'qikB'],\n",
       "       [2008, -0.6265848786348389, -2.214012989, ..., 295.98, 5, 'qikB']],\n",
       "      dtype=object)"
      ]
     },
     "metadata": {},
     "execution_count": 162
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "source": [
    "# dataset[row indices, column indices]\n",
    "X =  dataset[:,(0,2,3,4,5)] # input\n",
    "Y = dataset[:, 1] # output"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "source": [
    "X"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([[1967, -0.13204552, 73.25, 274.75, 5],\n",
       "       [1968, -0.13489132, 73.25, 274.75, 5],\n",
       "       [1969, -0.13773713, 73.25, 274.75, 5],\n",
       "       ...,\n",
       "       [2006, -2.200303795, 67.56, 295.98, 5],\n",
       "       [2007, -2.1436467009999998, 67.56, 295.98, 5],\n",
       "       [2008, -2.214012989, 67.56, 295.98, 5]], dtype=object)"
      ]
     },
     "metadata": {},
     "execution_count": 164
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "source": [
    "Y"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([-1.3019167168272867, -0.1755116544922609, -0.5753912617866366,\n",
       "       -1.3146528982453876, -1.3448122739791868, -1.3920445226960712,\n",
       "       -1.3359799310565001, -1.3849985450506213, -1.1754808111323247,\n",
       "       -1.211653700305356, -0.013779809801942856, -1.0094600215864684,\n",
       "       -1.3693388253450396, -0.870893515991459, -0.9965212498274112,\n",
       "       -1.2972207400533886, -1.4433993001778924, -1.3935131538245413,\n",
       "       -0.9026492390936863, -1.186396862069766, -1.2970876296361284,\n",
       "       -1.2265439720617402, -1.2031206612785652, -1.2614099631706877,\n",
       "       -1.2190537187788226, -1.2732156051529773, -1.255806233319971,\n",
       "       -1.1233683874209723, -0.7148553392520015, -1.0238563213497403,\n",
       "       -0.7783196163363754, -0.403887448521952, -0.48413105389206756,\n",
       "       -0.6506048443778936, -0.8245940825177562, -1.0313526487184896,\n",
       "       -1.1179656643006537, -1.1998982553680737, -0.4817744810522229,\n",
       "       0.044552326532235904, 0.18978038520112425, -0.2639748965512908,\n",
       "       5.366422533988953, 5.025528419900824, 5.311603702880718,\n",
       "       5.494605170355903, 5.521315526079248, 4.959700836075677,\n",
       "       5.1391611033015785, 5.331595838069916, 5.489191677835253,\n",
       "       5.126178750285396, 4.766005750055666, 5.854147560066647,\n",
       "       5.243788767743994, 5.680414619269194, 5.720472746425205,\n",
       "       5.973358545038434, 5.433487576467019, 5.0862339293515255,\n",
       "       5.414406189212093, 5.083994549733621, 5.659065549020415,\n",
       "       5.288176876527291, 4.75548071552206, 5.309942243275819,\n",
       "       5.507048375076717, 5.345857679843903, 5.106305647779394,\n",
       "       5.67532198517411, 5.664351531752834, 5.217834874435709,\n",
       "       6.266719893172937, 6.386353057843667, 5.313811717209993,\n",
       "       5.038602341104437, 5.570131623082691, 5.043735749191708,\n",
       "       5.742397372369413, 6.325336902229878, 5.531250139077504,\n",
       "       5.6656775662192596, 5.098340921931797, 5.649839992876407,\n",
       "       5.806004680969097, 5.17607444635144, 5.353927323111781,\n",
       "       5.455051954145785, 5.720489199514742, 5.780040498133059,\n",
       "       4.924583004580604, 5.506970818395971, 5.570067719176963,\n",
       "       5.617996794206124, 5.358670278831765, 6.149005280600654,\n",
       "       5.600888592225534, 5.295247027167568, 5.809916686128686,\n",
       "       5.444500925364317, 5.983448198548069, 5.479221558129346,\n",
       "       5.1707913323684975, 5.581028772725, 5.851154168446858,\n",
       "       5.753206052162029, 5.564680238564809, 6.004171824013746,\n",
       "       5.970585240258111, 5.820041371716393, 6.714153916747481,\n",
       "       6.345831363289445, 5.615210316799305, 5.456193994592737,\n",
       "       5.736425898693227, 5.125501319214148, 5.5356769174061435,\n",
       "       6.206029468112523, 1.3284102900005466, 1.8064630837179716,\n",
       "       1.6518569005291084, 3.1419712654703, 1.9846162315871991,\n",
       "       1.2406354267150168, 1.4052977120089867, 2.411525293408583,\n",
       "       1.8811393718545635, 3.6691153614471363, 2.551034466673931,\n",
       "       2.213602214391965, 1.7163564117913164, 1.8928663116983242,\n",
       "       2.8192658948505085, 3.4085927949183508, 1.775262243010932,\n",
       "       2.124964486548884, 0.6625701760252315, 1.3144118209472961,\n",
       "       1.673074927594927, 2.3727272516633895, 2.1648948334364424,\n",
       "       1.8482899553039007, 1.461971643574846, 1.6463464099991443,\n",
       "       1.2802519906933107, 1.516608748688466, 2.053302783978224,\n",
       "       1.2332686590444706, 2.0385011115835776, 2.704040887144705,\n",
       "       3.7456833801870726, 2.1821476503068373, 2.158749379165885,\n",
       "       2.194620533261008, 2.560939185075565, 2.958709852281674,\n",
       "       1.8018566776687897, 1.5248316423967483, 2.146708284401231,\n",
       "       1.5956877101626663, 1.8278395582197449, 1.6723406645986765,\n",
       "       0.6648230521241207, 2.0586075584093733, 1.697061650454998,\n",
       "       1.237677087981461, 1.473008658840425, 0.5946603549204563,\n",
       "       2.137875333428382, -1.2155561972675581, -1.286791174716893,\n",
       "       -1.1727307428428462, -0.5923384893242087, -1.123516489388532,\n",
       "       -1.4315381149450939, -1.238402855538187, -1.0641709300024167,\n",
       "       -1.0174701171261926, -0.7092904315463137, -0.9500059733233815,\n",
       "       -1.3592881684501972, -0.8645386666591678, -0.863454568953741,\n",
       "       -1.4493820440201526, -1.3141742884403183, -1.1783122425632815,\n",
       "       -1.0816000014366136, -0.8375872004599801, -0.8167541943978339,\n",
       "       -1.1031639109409992, -1.2674364188597316, -0.8203187590198859,\n",
       "       -0.8500448409911419, -0.4846205130396852, -0.4231240856017745,\n",
       "       -0.5300163786326134, -0.9363431642829836, -0.6086970386317088,\n",
       "       -0.5753716832093357, -1.1384477413126397, -0.9038943260176375,\n",
       "       -1.3683738971040358, -1.0314118970030297, -0.294161577132486,\n",
       "       -0.34998577832898486, -0.6265848786348389, -0.9500059733233815,\n",
       "       -1.3592881684501972, -0.8645386666591678, -0.863454568953741,\n",
       "       -1.4493820440201526, -1.3141742884403183, -1.1783122425632815,\n",
       "       -1.0816000014366136, -0.8375872004599801, -0.8167541943978339,\n",
       "       -1.1031639109409992, -1.2674364188597316, -0.8203187590198859,\n",
       "       -0.8500448409911419, -0.4846205130396852, -0.4231240856017745,\n",
       "       -0.5300163786326134, -0.9363431642829836, -0.6086970386317088,\n",
       "       -0.5753716832093357, -1.1384477413126397, -0.9038943260176375,\n",
       "       -1.3683738971040358, -1.0314118970030297, -0.294161577132486,\n",
       "       -0.34998577832898486, -0.6265848786348389], dtype=object)"
      ]
     },
     "metadata": {},
     "execution_count": 165
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "source": [
    "from sklearn import preprocessing"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "source": [
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "X_scale = min_max_scaler.fit_transform(X)\n",
    "X_scale"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([[0.18      , 0.88054431, 1.        , 0.7999685 , 0.        ],\n",
       "       [0.2       , 0.87961982, 1.        , 0.7999685 , 0.        ],\n",
       "       [0.22      , 0.87869532, 1.        , 0.7999685 , 0.        ],\n",
       "       ...,\n",
       "       [0.96      , 0.20864845, 0.73934952, 0.96716018, 0.        ],\n",
       "       [0.98      , 0.22705412, 0.73934952, 0.96716018, 0.        ],\n",
       "       [1.        , 0.20419487, 0.73934952, 0.96716018, 0.        ]])"
      ]
     },
     "metadata": {},
     "execution_count": 167
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "source": [
    "from sklearn.model_selection import train_test_split"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "source": [
    "X_train, X_val_and_test, Y_train, Y_val_and_test = train_test_split(\n",
    "    X_scale, Y, test_size=0.2)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "source": [
    "X_train.shape"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(186, 5)"
      ]
     },
     "metadata": {},
     "execution_count": 170
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "source": [
    "X_val_and_test.shape"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(47, 5)"
      ]
     },
     "metadata": {},
     "execution_count": 171
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers.advanced_activations import LeakyReLU"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "source": [
    "model = Sequential([\n",
    "    Dense(32, activation=tf.keras.layers.LeakyReLU(alpha=0.2), input_shape=(5,), kernel_initializer='normal'),\n",
    "    #kernel_initializer='normal'\n",
    "#    Dense(32, activation='relu', input_shape=(5,)),\n",
    "#      Dense(32, activation='relu'),\n",
    "#      Dense(16, activation='relu'),\n",
    "#      Dense(16, activation='relu'),\n",
    "#      Dense(16, activation='relu'),\n",
    "    Dense(1, activation=tf.keras.layers.LeakyReLU(alpha=0.2)),\n",
    "])"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "source": [
    "model.compile(\n",
    "    optimizer='adam',\n",
    "    loss='mean_squared_error',\n",
    "    metrics=[\n",
    "        'MeanSquaredError',\n",
    "    ]\n",
    ")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "source": [
    "# from sklearn.metrics import r2_score\n",
    "\n",
    "# model.compile(optimizer='sgd',\n",
    "#               loss='binary_crossentropy',\n",
    "#               metrics=['r2_score'])"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "source": [
    "X_train = np.asarray(X_train).astype(np.float32)\n",
    "Y_train = np.asarray(Y_train).astype(np.float32)\n",
    "X_val_and_test = np.asarray(X_val_and_test).astype(np.float32)\n",
    "Y_val_and_test = np.asarray(Y_val_and_test).astype(np.float32)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "source": [
    "Y_val_and_test"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([ 5.736426  , -0.9363432 ,  5.7532063 , -0.5753717 , -0.7148553 ,\n",
       "        5.7423973 ,  2.1378753 , -0.34998578,  3.7456834 ,  3.4085927 ,\n",
       "       -0.95000595, -1.00946   ,  5.0255284 ,  5.3458576 ,  5.8511543 ,\n",
       "        2.4115252 ,  6.714154  ,  2.560939  ,  3.1419713 , -0.8375872 ,\n",
       "        1.8018566 , -0.70929044,  5.8541474 , -1.103164  ,  1.673075  ,\n",
       "       -0.87089354,  5.570132  , -1.2031207 ,  4.755481  ,  1.2376771 ,\n",
       "       -1.1233684 ,  2.1467083 , -1.3448123 ,  5.53125   , -0.01377981,\n",
       "        1.5956877 ,  5.5070486 , -0.53001636,  6.2060294 ,  1.7163564 ,\n",
       "        0.18978038,  5.6656775 ,  5.455052  ,  1.84829   ,  5.043736  ,\n",
       "       -1.1863968 , -1.2674364 ], dtype=float32)"
      ]
     },
     "metadata": {},
     "execution_count": 177
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "source": [
    "hist = model.fit(X_train, Y_train,\n",
    "          batch_size=16, epochs=64,\n",
    "          validation_data=(X_val_and_test, Y_val_and_test))"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Epoch 1/64\n",
      "12/12 [==============================] - 0s 10ms/step - loss: 10.8651 - mean_squared_error: 10.8651 - val_loss: 12.0388 - val_mean_squared_error: 12.0388\n",
      "Epoch 2/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 10.5149 - mean_squared_error: 10.5149 - val_loss: 11.6091 - val_mean_squared_error: 11.6091\n",
      "Epoch 3/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 10.1826 - mean_squared_error: 10.1826 - val_loss: 11.1804 - val_mean_squared_error: 11.1804\n",
      "Epoch 4/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 9.8434 - mean_squared_error: 9.8434 - val_loss: 10.7271 - val_mean_squared_error: 10.7271\n",
      "Epoch 5/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 9.4542 - mean_squared_error: 9.4542 - val_loss: 10.2356 - val_mean_squared_error: 10.2356\n",
      "Epoch 6/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 9.0251 - mean_squared_error: 9.0251 - val_loss: 9.7106 - val_mean_squared_error: 9.7106\n",
      "Epoch 7/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 8.5859 - mean_squared_error: 8.5859 - val_loss: 9.1186 - val_mean_squared_error: 9.1186\n",
      "Epoch 8/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 8.0802 - mean_squared_error: 8.0802 - val_loss: 8.5287 - val_mean_squared_error: 8.5287\n",
      "Epoch 9/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 7.5821 - mean_squared_error: 7.5821 - val_loss: 7.8955 - val_mean_squared_error: 7.8955\n",
      "Epoch 10/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 7.0510 - mean_squared_error: 7.0510 - val_loss: 7.2833 - val_mean_squared_error: 7.2833\n",
      "Epoch 11/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 6.5332 - mean_squared_error: 6.5332 - val_loss: 6.6542 - val_mean_squared_error: 6.6542\n",
      "Epoch 12/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 6.0096 - mean_squared_error: 6.0096 - val_loss: 6.0556 - val_mean_squared_error: 6.0556\n",
      "Epoch 13/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 5.5237 - mean_squared_error: 5.5237 - val_loss: 5.4727 - val_mean_squared_error: 5.4727\n",
      "Epoch 14/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 5.0105 - mean_squared_error: 5.0105 - val_loss: 4.9315 - val_mean_squared_error: 4.9315\n",
      "Epoch 15/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 4.5238 - mean_squared_error: 4.5238 - val_loss: 4.4333 - val_mean_squared_error: 4.4333\n",
      "Epoch 16/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 4.0553 - mean_squared_error: 4.0553 - val_loss: 3.9659 - val_mean_squared_error: 3.9659\n",
      "Epoch 17/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 3.6226 - mean_squared_error: 3.6226 - val_loss: 3.4996 - val_mean_squared_error: 3.4996\n",
      "Epoch 18/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 3.2119 - mean_squared_error: 3.2119 - val_loss: 3.0531 - val_mean_squared_error: 3.0531\n",
      "Epoch 19/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 2.8122 - mean_squared_error: 2.8122 - val_loss: 2.6555 - val_mean_squared_error: 2.6555\n",
      "Epoch 20/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 2.4511 - mean_squared_error: 2.4511 - val_loss: 2.3060 - val_mean_squared_error: 2.3060\n",
      "Epoch 21/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 2.1254 - mean_squared_error: 2.1254 - val_loss: 2.0220 - val_mean_squared_error: 2.0220\n",
      "Epoch 22/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 1.8484 - mean_squared_error: 1.8484 - val_loss: 1.7516 - val_mean_squared_error: 1.7516\n",
      "Epoch 23/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 1.6037 - mean_squared_error: 1.6037 - val_loss: 1.5315 - val_mean_squared_error: 1.5315\n",
      "Epoch 24/64\n",
      "12/12 [==============================] - 0s 4ms/step - loss: 1.3892 - mean_squared_error: 1.3892 - val_loss: 1.3601 - val_mean_squared_error: 1.3601\n",
      "Epoch 25/64\n",
      "12/12 [==============================] - 0s 5ms/step - loss: 1.2142 - mean_squared_error: 1.2142 - val_loss: 1.2125 - val_mean_squared_error: 1.2125\n",
      "Epoch 26/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 1.0651 - mean_squared_error: 1.0651 - val_loss: 1.0762 - val_mean_squared_error: 1.0762\n",
      "Epoch 27/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.9471 - mean_squared_error: 0.9471 - val_loss: 0.9892 - val_mean_squared_error: 0.9892\n",
      "Epoch 28/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.8493 - mean_squared_error: 0.8493 - val_loss: 0.9191 - val_mean_squared_error: 0.9191\n",
      "Epoch 29/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.7705 - mean_squared_error: 0.7705 - val_loss: 0.8635 - val_mean_squared_error: 0.8635\n",
      "Epoch 30/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 0.7096 - mean_squared_error: 0.7096 - val_loss: 0.8262 - val_mean_squared_error: 0.8262\n",
      "Epoch 31/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.6532 - mean_squared_error: 0.6532 - val_loss: 0.7832 - val_mean_squared_error: 0.7832\n",
      "Epoch 32/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.6101 - mean_squared_error: 0.6101 - val_loss: 0.7540 - val_mean_squared_error: 0.7540\n",
      "Epoch 33/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.5701 - mean_squared_error: 0.5701 - val_loss: 0.7296 - val_mean_squared_error: 0.7296\n",
      "Epoch 34/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.5372 - mean_squared_error: 0.5372 - val_loss: 0.7053 - val_mean_squared_error: 0.7053\n",
      "Epoch 35/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.5072 - mean_squared_error: 0.5072 - val_loss: 0.6826 - val_mean_squared_error: 0.6826\n",
      "Epoch 36/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.4810 - mean_squared_error: 0.4810 - val_loss: 0.6767 - val_mean_squared_error: 0.6767\n",
      "Epoch 37/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.4549 - mean_squared_error: 0.4549 - val_loss: 0.6457 - val_mean_squared_error: 0.6457\n",
      "Epoch 38/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.4331 - mean_squared_error: 0.4331 - val_loss: 0.6281 - val_mean_squared_error: 0.6281\n",
      "Epoch 39/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.4136 - mean_squared_error: 0.4136 - val_loss: 0.6109 - val_mean_squared_error: 0.6109\n",
      "Epoch 40/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 0.3955 - mean_squared_error: 0.3955 - val_loss: 0.5861 - val_mean_squared_error: 0.5861\n",
      "Epoch 41/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.3780 - mean_squared_error: 0.3780 - val_loss: 0.5844 - val_mean_squared_error: 0.5844\n",
      "Epoch 42/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.3621 - mean_squared_error: 0.3621 - val_loss: 0.5701 - val_mean_squared_error: 0.5701\n",
      "Epoch 43/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.3463 - mean_squared_error: 0.3463 - val_loss: 0.5583 - val_mean_squared_error: 0.5583\n",
      "Epoch 44/64\n",
      "12/12 [==============================] - 0s 3ms/step - loss: 0.3341 - mean_squared_error: 0.3341 - val_loss: 0.5423 - val_mean_squared_error: 0.5423\n",
      "Epoch 45/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.3220 - mean_squared_error: 0.3220 - val_loss: 0.5325 - val_mean_squared_error: 0.5325\n",
      "Epoch 46/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.3104 - mean_squared_error: 0.3104 - val_loss: 0.5168 - val_mean_squared_error: 0.5168\n",
      "Epoch 47/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.3001 - mean_squared_error: 0.3001 - val_loss: 0.5159 - val_mean_squared_error: 0.5159\n",
      "Epoch 48/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2929 - mean_squared_error: 0.2929 - val_loss: 0.5005 - val_mean_squared_error: 0.5005\n",
      "Epoch 49/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2825 - mean_squared_error: 0.2825 - val_loss: 0.5002 - val_mean_squared_error: 0.5002\n",
      "Epoch 50/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2758 - mean_squared_error: 0.2758 - val_loss: 0.4888 - val_mean_squared_error: 0.4888\n",
      "Epoch 51/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2694 - mean_squared_error: 0.2694 - val_loss: 0.4989 - val_mean_squared_error: 0.4989\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Epoch 52/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2627 - mean_squared_error: 0.2627 - val_loss: 0.4813 - val_mean_squared_error: 0.4813\n",
      "Epoch 53/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2572 - mean_squared_error: 0.2572 - val_loss: 0.4702 - val_mean_squared_error: 0.4702\n",
      "Epoch 54/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2521 - mean_squared_error: 0.2521 - val_loss: 0.4621 - val_mean_squared_error: 0.4621\n",
      "Epoch 55/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2488 - mean_squared_error: 0.2488 - val_loss: 0.4554 - val_mean_squared_error: 0.4554\n",
      "Epoch 56/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2452 - mean_squared_error: 0.2452 - val_loss: 0.4731 - val_mean_squared_error: 0.4731\n",
      "Epoch 57/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2412 - mean_squared_error: 0.2412 - val_loss: 0.4632 - val_mean_squared_error: 0.4632\n",
      "Epoch 58/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2361 - mean_squared_error: 0.2361 - val_loss: 0.4529 - val_mean_squared_error: 0.4529\n",
      "Epoch 59/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2357 - mean_squared_error: 0.2357 - val_loss: 0.4372 - val_mean_squared_error: 0.4372\n",
      "Epoch 60/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2315 - mean_squared_error: 0.2315 - val_loss: 0.4394 - val_mean_squared_error: 0.4394\n",
      "Epoch 61/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2281 - mean_squared_error: 0.2281 - val_loss: 0.4508 - val_mean_squared_error: 0.4508\n",
      "Epoch 62/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2276 - mean_squared_error: 0.2276 - val_loss: 0.4303 - val_mean_squared_error: 0.4303\n",
      "Epoch 63/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2249 - mean_squared_error: 0.2249 - val_loss: 0.4363 - val_mean_squared_error: 0.4363\n",
      "Epoch 64/64\n",
      "12/12 [==============================] - 0s 2ms/step - loss: 0.2259 - mean_squared_error: 0.2259 - val_loss: 0.4427 - val_mean_squared_error: 0.4427\n"
     ]
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "source": [
    "model.evaluate(X_val_and_test, Y_val_and_test)[1]"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "2/2 [==============================] - 0s 2ms/step - loss: 0.4427 - mean_squared_error: 0.4427\n"
     ]
    },
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "0.44266340136528015"
      ]
     },
     "metadata": {},
     "execution_count": 179
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "This Sequential Neural Network Model produced an MSE of 0.4427."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "source": [
    "import matplotlib.pyplot as plt"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "source": [
    "plt.plot(hist.history['mean_squared_error'])\n",
    "plt.plot(hist.history['val_mean_squared_error'])\n",
    "plt.title('Model RMSE')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper right')\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEWCAYAAABhffzLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAut0lEQVR4nO3dd5wV1f3/8dfnlu2N3pamUkRpuiqCBUWiUayJsX81MfGnRrFEsWs0xhhjxZJo1KjRqLFg70bEggUEFUTAQm9L34Utd/ee3x8zK8vSdmH3zt697+cj85i5c+feec8GPzNzZu4Zc84hIiKpIxR0ABERSSwVfhGRFKPCLyKSYlT4RURSjAq/iEiKUeEXEUkxKvwigJn1MDNnZpF6LHuGmX2YiFwiTUGFX5KOmc0xs0oza1tn/lS/ePcIKFrtHUipP8wxs8vrLFOv/GZWaGbPmdlyM1tjZl+b2RlbWE/NcEKitlWSlwq/JKsfgZNqXphZfyAzuDibKHDO5QC/BK4xs5F13q9P/n8D84HuQBvg/4Clm1tPreHpxtwIaZlU+CVZ/RuvENY4HXis9gJmlm9mj5lZsZnNNbOrzSzkvxc2s1v9o+kfgCM289mHzGyxmS00sxvNLNzQkM65ScB0YFBD8wN7AY8459Y556qcc1Occ683NINIXSr8kqw+AfLMbFe/IJ8APF5nmbuBfGAn4EC8Qvtr/73fAaOAwUAR3pF5bY8CVcAu/jI/A37b0JBmNgTYHfhuO/J/AtxrZieaWbeGrltkS1T4JZnVHDWPBL4FFta8UauYXuGcK3HOzQFuA07zF/kVcKdzbr5zbiXwl1qf7QD8HLjQP9peBtwBnNiAbMvNrAyYCNwHvNCQ/L7jgQ+Aa4Af/WsAe21mPatrDbs2IKOkqG3ewSDSjP0bmAD0ZNNmkrZAGjC31ry5QBd/ujNe+3nt92p0B6LAYjOrmReqs/y2tAUccCFeW34UqGxAfpxzq4DLgcv9C8G3Ai+YWWHt9TjnqhqQS0RH/JK8nHNz8S6SHg48X+ft5UAMr4jX6MaGo+rFQNc679WYD1TgFdUCf8hzzu3WwHzVzrnbgHLg3Abmr7vscrzC3xlo3ZAcInWp8EuyOxM42Dm3rvZM51w18F/gz2aWa2bdgYvZ0I7+X2C0f8tkK7wj65rPLgbeAm4zszwzC5nZzmZ24HZmvBkYY2YZ9c0PYGZ/NbPdzSxiZrnAOcB3zrkV25lDBFDhlyTnnPvev3Nmc84H1gE/AB8C/wEe9t/7J/Am8CXwBZsecf8fXlPRN8Aq4Fmg03bGfNX/jt81MH8WMA5Y7W9Dd+CoOsusrnMf/8XbmVFSiOlBLCIiqUVH/CIiKUaFX0Qkxajwi4ikGBV+EZEUkxQ/4Grbtq3r0aNH0DFERJLK5MmTlzvn2tWdnxSFv0ePHkyatKU73kREZHPMbO7m5qupR0Qkxajwi4ikGBV+EZEUkxRt/CIiDRWLxViwYAHl5eVBR2lyGRkZFBYWEo1G67W8Cr+ItEgLFiwgNzeXHj16UKt77RbHOceKFStYsGABPXv2rNdn1NQjIi1SeXk5bdq0adFFH8DMaNOmTYPObJqs8JvZw2a2zMym1Zr3NzP71sy+MrNxZlbQVOsXEWnpRb9GQ7ezKY/4HwEOqzPvbWB359wAYBZwRROuH+Z9Ch/e0aSrEBFJNk1W+J1zE4CVdea9VesxcZ8AhZt8sDFNHwfv/BF+/KBJVyMiUteKFSsYNGgQgwYNomPHjnTp0uWn15WVdZ/CubFJkyYxevToJssW5MXd3wBPb+lNMzsLOAugW7duW1ps60ZcC7PfhBfPhXMmQnrO9n2PiEgDtWnThqlTpwLwxz/+kZycHC655JKf3q+qqiIS2XwJLioqoqioqMmyBXJx18yuAqqAJ7a0jHPuAedckXOuqF27TbqaqJ+0LDjm77B6Prx97fZ9h4hIIznjjDO4+OKLOeigg7jsssv47LPPGDp0KIMHD2bo0KHMnDkTgPHjxzNq1CjA22n85je/Yfjw4ey0006MHTt2h3Mk/IjfzE4HRgEjXCIe/9VtCOz7e5h4D+x6JOx8UJOvUkSal+tfns43i9Y26nf265zHdUfu1uDPzZo1i3feeYdwOMzatWuZMGECkUiEd955hyuvvJLnnntuk898++23vPfee5SUlNCnTx/OOeecet+zvzkJLfxmdhhwGXCgc259wlZ88NUw60146Xw452PIyEvYqkVEajv++OMJh8MArFmzhtNPP53Zs2djZsRisc1+5ogjjiA9PZ309HTat2/P0qVLKSzc/kukTVb4zexJYDjQ1swWANfh3cWTDrzt3370iXPu7KbK8JNoptfk8/DP4K2r4agdP1USkeSxPUfmTSU7O/un6WuuuYaDDjqIcePGMWfOHIYPH77Zz6Snp/80HQ6Hqaqq2uxy9dVkhd85d9JmZj/UVOvbpq57wdDR8NGd0O8o2OWQwKKIiIB3xN+lSxcAHnnkkYStN7V+uTv8CmjXF148H8pWB51GRFLcmDFjuOKKKxg2bBjV1dUJW68l4vrqjioqKnKN9iCWhZPhwZEw8EQ45r7G+U4RaXZmzJjBrrvuGnSMhNnc9prZZOfcJveFptYRP0CXPWG/i2DqEzDzjaDTiIgkXOoVfoADL4MOu8PLo2H9ym0vLyLSgqRm4Y+kec0861fA65cFnUZEJKFSs/ADdBoIB1wKX/8XZrwcdBoRkYRJ3cIPsP8foOMAeOUiWLci6DQiIgmR2oU/HIVj/+Hd2vn6pUGnERFJCD16scNucMAlMP4vMOBE6P2zoBOJSAuwYsUKRowYAcCSJUsIh8PUdDj52WefkZaWttXPjx8/nrS0NIYOHdro2VT4wbu9c/o4ePVi6P6Jum8WkR22rW6Zt2X8+PHk5OQ0SeFP7aaeGpF0OHIsrFkA/7sx6DQi0kJNnjyZAw88kD333JNDDz2UxYsXAzB27Fj69evHgAEDOPHEE5kzZw7/+Mc/uOOOOxg0aBAffNC4D5Nq8Uf85bFqMqLhbS/YbR/Y60z49B/Q/3go3LPpw4lIYrx+OSz5unG/s2N/+PnN9V7cOcf555/Piy++SLt27Xj66ae56qqrePjhh7n55pv58ccfSU9PZ/Xq1RQUFHD22Wc3+Cyhvlr0Ef8db8/ihPsnUh6rZx8YI66D3E5e983Vm+8eVURke1RUVDBt2jRGjhzJoEGDuPHGG1mwYAEAAwYM4JRTTuHxxx/f4lO5GlOLPuLfvUs+Y/83mzHPfsVdJw7a9pPoM/LgiFvhqZPh47He7Z4ikvwacGTeVJxz7LbbbkycOHGT91599VUmTJjASy+9xJ/+9CemT5/epFla9BH/yH4duORnfXjpy0XcN/77+n2o7xHQ72gY/1dYUc/PiIhsQ3p6OsXFxT8V/lgsxvTp04nH48yfP5+DDjqIW265hdWrV1NaWkpubi4lJSVNkqVFF36Ac4fvzNGDOnPrWzN5+5ul9fvQz2/xLvi+fhkkQe+lItL8hUIhnn32WS677DIGDhzIoEGD+Pjjj6murubUU0+lf//+DB48mIsuuoiCggKOPPJIxo0b1yQXd1OiW+byWDUn3D+R75aV8ty5Q+nbsR6PXpx4L7x5JZz0NPQ5bLvXLSLBULfMKd4tc0Y0zP2nFZGdHuG3j05i5brKbX9o77OgbR9443KIlTd9SBGRBEmJwg/QMT+DB/6viGUlFZz978lUVG3jTp9w1LsgtOpH+OTexIQUEUmAlCn8AIO6FnDr8QP5bM5KLn3mK+LxbTRz7Xww9B0FE26FNQsTE1JEGk0yNGU3hoZuZ0oVfoCjBnZmzGHenT63vT1z2x849CZwcXj72qYPJyKNJiMjgxUrVrT44u+cY8WKFWRkZNT7My36Pv4tOefAnZm/cj33vvc9XVtlceLe3ba8cKvuMOwCeP+vUPQb6DEscUFFZLsVFhayYMECiouLg47S5DIyMigsLKz38ilZ+M2MPx29O4tWl3PVC9PoVJDJgb3bbfkDwy6EKU/A62PgrPchnJJ/NpGkEo1G6dmzZ9AxmqWUa+qpEQmHuPeUPejdIZdzH5/MN4vWbnnhtCw49EZYOg2mPp64kCIiTaDJCr+ZPWxmy8xsWq15rc3sbTOb7Y9bNdX66yMnPcK/ztiL3IwoZz76OcvWbuW2zX7HQNd94L2boKI0YRlFRBpbUx7xPwLU/eXT5cC7zrlewLv+60B1zM/goTOKWFMW48xHJ7G+smrzC5rByD9B6VKYeE9iQ4qINKImK/zOuQnAyjqzjwYe9acfBY5pqvU3xG6d87n7pMFMX7SGi56euuXbPLvtA7seBR+NhZJ6dv8gItLMJLqNv4NzbjGAP26/pQXN7Cwzm2RmkxJxVX7Erh24+oh+vDl9KX9989stL3jIH6G6wntUo4hIEmq2F3edcw8454qcc0U1z6lsar8e1oPThnTn/vd/4KnP5m1+oTY7Q9GZ8MVjUFyP3wGIiDQziS78S82sE4A/Xpbg9W+VmXHdkf04sHc7rn5hGhO/X7H5BQ8cA2nZ8M4fE5pPRKQxJLrwvwSc7k+fDryY4PVvUyQc4p6TB9OtTRYXPDWF5aUVmy6U3Rb2uxBmvgZzPkp4RhGRHdGUt3M+CUwE+pjZAjM7E7gZGGlms4GR/utmJzcjyr0n78HqshgX//fLzV/sHXIu5HWBt65Wn/0iklSa8q6ek5xznZxzUedcoXPuIefcCufcCOdcL39c966fZmPXTnlcd2Q/Jswq5v4JP2y6QDQTDroSFn0Bs95IfEARke3UbC/uNgcn792NIwZ04ta3ZjJ57mb2UQNOgILu8P4tOuoXkaShwr8VZsZfjutPl4JMzv/PFFavr/MAl3AU9r/YO+r//t1gQoqINJAK/zbkZUS55+TBFJdWcMkzX23axevAkyGvEN7/m476RSQpqPDXw4DCAi7/+a68M2Mp/6l7f38kzbvDZ/4nMKdxH4gsItIUVPjr6TfDerDfLm3586szmL9y/cZvDj4Ncjp6bf0iIs2cCn89mRk3/6I/ITPGPFvnsY3RDBg22jvin/dJcCFFROpBhb8BCltlcfURuzLxhxX8+5O5G7+5568hq62O+kWk2VPhb6AT9urKgb3bcfPr3zJn+boNb6RlwdDzvLt7Fk4OLqCIyDao8DdQTZNPJGxc+mydX/Xu9VvIKIAJtwaWT0RkW1T4t0On/EyuO3I3Pp+zioc/+nHDG+m5XlcOM1+DxV8FF1BEZCtU+LfTL/bowoi+7fnbmzP5objWoxj3+X+QngcT/hZcOBGRrVDh3041v+pNj4S47Llad/lkFnjFf8ZLsGxGoBlFRDZHhX8HtM/L4JpR/fh8zqqN7/IZci6k5aitX0SaJRX+HfTLPQs5sHc7/vrGtxt+2JXV2rvQO+05WD472IAiInWo8O8gM+Om47wfdl32XK2+fPY9DyIZ8MFtwQYUEalDhb8RdCnI5IrD+/Lx9yt46vP53sycdlD0G/jqv7ByM/35i4gERIW/kZy8dzeG7tyGP786g0Wry7yZw0ZDKAIf3B5sOBGRWlT4G4mZcfNxA6iOO64a97XX5JPbEfY8Hb58ElbP2/aXiIgkgAp/I+rWJos//Kw3780s5r2Zy7yZwy4ADD68M8hoIiI/UeFvZKcP7cFO7bK58ZUZxKrjkF8Ig06GqU9AaXHQ8UREVPgbWzQc4uojduWH5ev490T/3v6h50NVOXz+z2DDiYigwt8kDurTnv17teXOd2axal0ltO0FfQ6Hz/4Jleu3/QUiIk1Ihb8JmBnXjOpHaUUVd74zy5s5dDSUrfSafEREAqTC30R6d8jllH268/in85i1tAS6DYHCvWDiPRCvDjqeiKQwFf4mdNHI3mSlhfnTK9/gwGvrXzUHZrwccDIRSWWBFH4zu8jMppvZNDN70swygsjR1Fpnp3HBiF58MHs542cWQ99R0Hon+HgsOLftLxARaQIJL/xm1gUYDRQ553YHwsCJic6RKP+3bw92apvNn175hsq4wb6/9x7NOG9i0NFEJEUF1dQTATLNLAJkAYsCytHk0iIhrh7l3d752MQ5MPBkyGoDH40NOpqIpKiEF37n3ELgVmAesBhY45x7q+5yZnaWmU0ys0nFxcn9w6eD+rTnwN7tuOud2SyvDMNev4NZr0PxzKCjiUgKCqKppxVwNNAT6Axkm9mpdZdzzj3gnCtyzhW1a9cu0TEbVc3tnWWxam57aybs/Tuvy+aPddQvIokXRFPPIcCPzrli51wMeB4YGkCOhNqlfQ6nD+3BU5/PZ9rqKAw+Db58GtYsDDqaiKSYIAr/PGCImWWZmQEjgJR4OO3oEb1onZXG9S9Pxw09D1wcJt4bdCwRSTFBtPF/CjwLfAF87Wd4INE5gpCfGeWSQ/vw+ZxVvDIvDfofD5MfgfUrg44mIikkkLt6nHPXOef6Oud2d86d5pyrCCJHEH5V1JXdOufxl9dmUL7P+RBbB5/eH3QsEUkh+uVugoVDxnVH7saiNeX8Y0aa33nb/VBRGnQ0EUkRKvwB2Ltnaw7v35EHJvzAyj1+D2Wr4ItHg44lIilChT8glx7al8qqOLfPyIce+8PH90BVyrR4iUiAVPgD0rNtNifv040nP5vPov7nQMki+OrpoGOJSApQ4Q/Q6BG9yIiEuGF6B+g4wHsur7psFpEmpsIfoLY56Zx1wM688c1Svu/7/2Dl9+qyWUSanAp/wH67f0/a5qRz5YzuuNY7wUd3qctmEWlSKvwBy06PcOEhvfh07lq+7XE6LPoC5n4UdCwRacFU+JuBE/bqyk7tsvnD7H64rLbeUb+ISBNR4W8GouEQYw7tyzfFMb7ucgLMfguWfhN0LBFpoVT4m4lDd+tAUfdWXPTDXrhoFnx8d9CRRKSFUuFvJsyMq47Yle/XpTG1zSj4+hl12SwiTUKFvxkZ3K0VRw7szCUL98O5OHz696AjiUgLpMLfzIw5tA/zXXum5g6HSY9A+ZqgI4lIC6PC38x0bZ3Fr4f14JrlB0NlCUz6V9CRRKSFUeFvhs4dvgsLM3rzddog3Kf/UOdtItKoVPibofzMKBce0ptbSg/DShbDl08FHUlEWpB6FX4zyzazkD/d28yOMrNo00ZLbSfv042FrYcwM7Qz7sM7oLoq6Egi0kLU94h/ApBhZl2Ad4FfA480VSjxftR1xeH9uL38KGzVjzD9+aAjiUgLUd/Cb8659cBxwN3OuWOBfk0XSwAO2bU9q7uN5Du6Uj3hVojHg44kIi1AvQu/me0LnAK86s+LNE0kqWFmXHnEboytPIrw8pkw89Vtf0hEZBvqW/gvBK4AxjnnppvZTsB7TZZKfjKwawG2+7HMcR2JvXeLumwWkR1Wr8LvnHvfOXeUc+6v/kXe5c650U2cTXyXHLYbD1QfRXTZV/Ddu0HHEZEkV9+7ev5jZnlmlg18A8w0s0ubNprU6No6i9x9TmWha8P6d2/WUb+I7JD6NvX0c86tBY4BXgO6Aadt70rNrMDMnjWzb81shn/9QLbi3BH9eMyOJmvJ53pQi4jskPoW/qh/3/4xwIvOuRiwI4eddwFvOOf6AgOBGTvwXSkhPytKx+G/o9jls+qNm4KOIyJJrL6F/35gDpANTDCz7sDa7VmhmeUBBwAPATjnKp1zq7fnu1LNyfv14Zm0o2m15COq534adBwRSVL1vbg71jnXxTl3uPPMBQ7aznXuBBQD/zKzKWb2oH/tYCNmdpaZTTKzScXFxdu5qpYlPRKm589Hs8LlsvTl64OOIyJJqr4Xd/PN7PaaQmxmt+Ed/W+PCLAH8Hfn3GBgHXB53YWccw8454qcc0Xt2rXbzlW1PIcN3oU3835J5+Ufsfa7T4KOIyJJqL5NPQ8DJcCv/GEtsL39BS8AFjjnatoqnsXbEUg9mBlFx1/GKpfDohd11C8iDVffwr+zc+4659wP/nA9XpNNgznnlgDzzayPP2sE3i2iUk+9u3ViSpdT6FvyMd9NmRB0HBFJMvUt/GVmtl/NCzMbBpTtwHrPB54ws6+AQYBuU2mgohMuZw05rHz9RuJx3dcvIvVX3/52zgYeM7N8//Uq4PTtXalzbipQtL2fF8jLb820Pr9m75l38/Z7bzFyxKFBRxKRJFHfu3q+dM4NBAYAA/yLsgc3aTLZpn5HX0Kp5RD94G+sKYsFHUdEkkSDnsDlnFvr/4IX4OImyCMNEMoqYN3g3zGcz3nyxVeCjiMiSWJHHr1ojZZCtluHkRdSFsqh5zf38t2ykqDjiEgS2JHCryuKzUFmAfEh53Bo6HP+M+6FoNOISBLYauE3sxIzW7uZoQTonKCMsg3ZB4ymLFrAwQv+zgez9StnEdm6rRZ+51yucy5vM0Ouc05P4GouMvKIDB/DfuHpvPbCk1Tr9k4R2YodaeqRZiS6z29Zn9WZk0oe5tlJc4OOIyLNmAp/SxFJJ/Nn1zAg9CNT3nyM0oqqoBOJSDOlwt+C2IATKGvVh7NiT/DP92YGHUdEmikV/pYkFCbzsOvZKbSEVR89zKLVO9Krhoi0VCr8LU3vw6jovDfnhZ7j9temBp1GRJohFf6Wxoz0Q2+gva2i3fRH+OzHlUEnEpFmRoW/Jeq+L9W7HMp50Re5fdwHVFXHg04kIs2ICn8LFT7sJjItxrErH+Y/n80LOo6INCMq/C1V212wIWdzfOR9XnvzNVaUVgSdSESaCRX+FswOHEM8sw2Xxv/F3974Nug4ItJMqPC3ZBn5REZex56hWZRNeZqp81cHnUhEmgEV/pZu0KlUdxzIldEnuemFSXpMo4io8Ld4oRDhw2+hAyvZb+m/+e+k+UEnEpGAqfCngm5DcP2P5+zIqzz2+gRWrqsMOpGIBEiFP0XYIdcTiUS4qPpf/OW1GUHHEZEAqfCnivwuhIZfxsjQJFZNeVG/6BVJYSr8qWTf84i37cuNaY9x47jPqazSL3pFUpEKfyoJRwkdeScdKebwlY/x0Ic/Bp1IRAKgwp9quu8Lg0/jd5HXeP3dd5i/cn3QiUQkwQIr/GYWNrMpZvZKUBlS1sgbIKOAP4Ye5PoXv8Y53dsvkkqCPOK/ANDtJUHIak340BvZw2bR5rtneGPakqATiUgCBVL4zawQOAJ4MIj1CzDoZOLdhnJ12pPc/sJHurdfJIUEdcR/JzAG2OJtJWZ2lplNMrNJxcXFCQuWMswIjbqDHKvgnNgjXPfS9KATiUiCJLzwm9koYJlzbvLWlnPOPeCcK3LOFbVr1y5B6VJM+77YsNEcF/qA5V+/zRvTFgedSEQSIIgj/mHAUWY2B3gKONjMHg8ghwDsfwmuoDu3ZDzK9eOmqMlHJAUkvPA7565wzhU653oAJwL/c86dmugc4kvLwo64na7xBfyq8nmufXFa0IlEpInpPn6BXodAv2M4P/IiX389hde/VpOPSEsWaOF3zo13zo0KMoP4DruZcDSdO3L+zdXjvtajGkVaMB3xiyevEzbiGvaITeGAyg+4/Hn9sEukpVLhlw32+i10Hsyfs57g029+4OnP9dAWkZZIhV82CIVh1J1kxlZxZ5vnuf7lb/hx+bqgU4lII1Phl411HoTtex4Hr3udYeFvuPDpqcSq1X2zSEuiwi+bGn4FtN6Ju7L/xcz5S7nnf98FnUhEGpEKv2wqLQuOHEv2unn8vcub3PPed0yeuyroVCLSSFT4ZfN67g97nsHwlf9leO4CLnp6KqUVVUGnEpFGoMIvWzbyBiynA3dnPcjSVWu59gX9qlekJVDhly3LyIcjbidr1Uwe7vUxz09ZyAtTFgadSkR2kAq/bF3fw2G3Yxm68GF+2XklV78wjXkr9LhGkWSmwi/bdvhtWGZr/uLuJMvKGf3UFN3iKZLEVPhl27LbwHEPEF31Pc/2eImp81dz5zuzgk4lIttJhV/qZ6cDYb+L6DbnWW7sNYv7xn/Px98vDzqViGwHFX6pv4OuhC5FnLLsdoa0KuXCp6aybG150KlEpIFU+KX+wlH45UMY8FDu/awvr+DcJ76gskrt/SLJRIVfGqZVDxh1B1lLJ/N8v/eZNHcV17+sB7WLJBMVfmm4/r+EwafSe+b93Lb7PJ74dB5PfTYv6FQiUk8q/LJ9Dr8NuuzJcXNv4JTua7j2xel8MU/9+YgkAxV+2T7RDDjxP1hGAX8q+zN98so45/HJLCvRxV6R5k6FX7Zfbkc46T+E1q/g6fz7KCsr46zHJlNWWR10MhHZChV+2TGdB8Mx95G15HNe2/l5vlywivOfnEJ1XM/rFWmuVPhlx+1+HBwwhsI5z/PM7p/zzoylXPfSND2sXaSZUuGXxjH8CtjtWIpm38mDvT/h8U/mcd/474NOJSKbEQk6gLQQoRAc909wjkO+GcvdXcs5/03olJ/BcXsUBp1ORGpJeOE3s67AY0BHIA484Jy7K9E5pAmEo/CLhyAU5shpD7C+fQVjnjVaZ6cxvE/7oNOJiC+II/4q4A/OuS/MLBeYbGZvO+e+CSCLNLZwBI59AEIRTvjqUWJ5Mc56zLj3lD0Y2a9D0OlEhAAKv3NuMbDYny4xsxlAF0CFv6UIR+CYv0MowqlTnyA9r4pzHo9zxwmDOXJg56DTiaS8QNv4zawHMBj4dDPvnQWcBdCtW7fEBpMdFwrDUfdAOMrxkx8hL3895z4Vp6Iqzi/3VJu/SJACK/xmlgM8B1zonFtb933n3APAAwBFRUW6LzAZhUIw6k5Iz+PQj8fyWKv1nP7MGZTFqjltSPeg04mkrEAKv5lF8Yr+E86554PIIAliBiNvgMwChr17A8+2LuOEF/4fK0srGT1iF8ws6IQiKSeIu3oMeAiY4Zy7PdHrlwCYwf5/gPQ8Br12Ca+2LuPYd87jx+Wl3PyLAWREw0EnFEkpQfyAaxhwGnCwmU31h8MDyCGJtvfv4Lh/snPZ13zQ6ga+/fITTn3wU1aUVgSdTCSlJLzwO+c+dM6Zc26Ac26QP7yW6BwSkAG/wk5/mYJwJa9mXkfvRS9wzH0fMXtpSdDJRFKGumyQxOsxDM7+gHD3fbgpfD+Xlt3Nyfe9xxvTlgSdTCQlqPBLMHLaw2kvwAFjOMr9j2ci13DfE//ljy9Np6JK3TqLNCUVfglOKAwHXwWnPEf3zDJeTL+W3p9dxa/vfZ05y9cFnU6kxVLhl+D1OgQ7bxK27+85MfoBf191Fo/ffTUvfTFXXTuLNAEVfmkeMvLg0D8TOucj0rvuwdX2ML1eGMVd941lno7+RRqVCr80L+37kvGbl6n+xSN0znZcWHwtq+7en3HPPEZ5ZVXQ6URaBBV+aX7MCPc/lvxLprBm5O0URks5dvr5zLp5P6aMH4eLx4NOKJLUVPil+QpHyR92Jm0un8Z3e91AJ7eMwePP4LubhvDte0+CdgAi20WFX5q/SBq7HHEB+ZdN49N+15BZtZq+75/NwpsGMufdf0J1LOiEIknFkuGuiaKiIjdp0qSgY0gzUV5RwcSXH6LLtH/Qm7mUWg5rC4fTvugYIr1HQmZB0BFFmgUzm+ycK9pkvgq/JKt15THee/VJmP48Q6on09bWUm1hqgqHkN77EOg+DDoPhkha0FFFAqHCLy1WVXWc92Ys4ZMP36LNgncZEZpCn9B8AFwkE+u6F3QbCl32gI4DILej12OoSAunwi8pYd6K9Tz1+Tw++HIGndd8yZDwtxycOZtuld9j+P/Ws9t5O4CO/aF1T2jVAwq6Q36h98B4kRZChV9SinOOrxas4ZWvFvHqV4tZu2Yl/ULzOKRgKftmL2Cnqh/IWjMbi9f6bYCFIb8LtNkF2vTyxzt7O4bcTpCWFdj2iGwPFX5JWfG448sFq/lg9nI+mF3MF/NWUx135EThwA6V7NumlP7Zq+kRXk5e2QJsxXew4nuorNNVdHq+10yU2wFyOnhnDlltvHF2243HaTlqTpLAqfCL+ErKY3zyw0o+/n45X85fzbRFa6ms8n4T0Do7jd0659G/cx57tI3RP72Y9tVLsNKlULIESpfA2sWwrhjWLd9051AjnO7vGFp7O4eaIbOVd7E5FPWalcJRiGZ5ZxR5nb0hPTeBfw1pyVT4RbagsirOzCUlTF2wmq/mr2b6orXMWlpCVdz7byM7LUzPdtn0bJtDz7bZ7Nwum+5tsunRJouCaLW3A1hXDOtX1Jpe7k2vX7FhWLcCKtZsO1BartdtdUYeZORDuj+uGdLzNrxX+4wjPU9nGbIRFX6RBqioqmbWklK+XriGWUtL+HH5On5YXsqCVWXU/k8mLyNC9zbZdGuTRWGrTDrnZ9IpP4POBZl0LsikVVZ04wfKx6u9H5zFY/64CipKvLOJtYtg7UIoWQyly6BiLZSv9cdrvOnYVjqsC6d5O4JwFDBvJ2Ah79pFWrbX/JSe409nQzTbu24RzfSns72zjfQ8f5zrvR/JhGiGNw5HtXNJIlsq/Al/2LpIMkiPhOlfmE//wvyN5pfHqpm3cj1zlq9j3sr1zF2xnrkr1zNt4Rrenr6UyuqNu5FIi4TokJdOx7wM2udl0DEvg7Y56bTNSaNtbjrtctJpm5NHm649iYbr8UP66pi3oyhfA+WrYf3KDWcZNWca8Wpwcbw9lPN2LpXrvGHtIqgs9aZjZd7YNeDBNxbydyA1OwZ/Z2LhWjsE85618NNyOd7OJJrpvQcblg1F/R1PptfkFc3wcleVe/mqyr0hkumf8dQ6C0rL9j4XyYSwSllD6K8l0gAZ0TC9O+TSu8Om7fDxuGP5ugoWry5n8ZoyFq0uZ+nacpas9cbfLFrL/2Ysoyy2+ULbKivq7xTSaZOTRuvsNAqy0miVFaV1dhr5mVHyM6MUZKWTn9mFvPzuROqzs9ga57ydSWwdVJR6O4WKkg1nG7EyqCqDWPmGYly5zru2UVEzlHo7j59OhWrtbGreryrbsZzbEvKvlUTSvOsrEX8Ip9UaIt44XrXhDKp8jbetLl7rzCbD36HUGUcza+3M8jacETnnfR5X6++53vtbxcq86VDYu75TewhFoLpyw5lfzXR15cavd/+Fd9txI1LhF2kkoZDRPjeD9rkZDOxasMXl1lVUsby0guWlFRSXVP407b2uYHlpJdMXrWXV+krWlMXYWmtsTnqEvIwIeZlRcjMi5GVEN5rO9d/LSY+QkxEh1x/npEfIzfDmhyNpXsHMbNX4f5Qa1TGvCALgNp1fUyBjZd7ZQO2iG8nwdhx1i3Xd4horg6oKr2BWlW+Yrvab1SrXQ3yNd3aSUQAF3TacPZj5Ozf/O2rONmJl3s5rXbG3jpqdWWz9trfZQl4TWjTT39ms9ncQDdRpoAq/SLLLTo+Qne5dG9iW6rhjTVmMlesqWVPm7QjWlMVYsz7G6rIYJeVVrCmLsbYsxtryGIvWlDNzaQkl5VWUlMeI1+MSXlZamBw/U2Y0TFZamMy0cK3piDeOevNrT9eMM6JhMiJhMtNCpEe812nhEGmRENGwEQ5FsIy8RvjrNRPVVd5ZT+U6/zpKrSEU9s4+wmkbXw+Jx70dVtkqb4hXb7izK5zmnQGE0zbMC0U3vG5kKvwizVg4ZLTO9pp9Gso5x7rKakrKY5SWV1FSUeWN/Z1CaUUV6yqqKa3wpksrqimrrGJ9ZTUl5VUsW1vB+lgVZZVxb36seqtnH1tjBmnhkLeDiIbIjHo7h/RIiGjYHyIh0sJGNBwiEvZ2GNFQiGjEm5cWrr2s9144ZETDRiQcIhIyImEjEqqZ9sahkBE2IxSCkNlG88KhOoM/LxL2x6GQ/53edE0dN0KQlo+lFxAyNr6AvyWhkNeBYGYB0LhH8A2lwi/SQpmZ18STHoH8bS+/Lc45KqrilFVWUxbzB3+6PFZNeSxOuT+/IlZNZbUjVh2nsipOrDpORVXcX27DshVVcaricWJVjrKyGJU1r/3PesPG09X1OY1JsHCdHUnIqDXtDRt2Gv7Y3xmF/fdDIcM5t6EhzHmNYjcf1599dmrTqHlV+EWkXszMP2IP04RXA7apOu7tCKrijip/Z1AVj1NV7aiKO6r9HUeVPz/uHNVxiDtHPO4v409X1wxuw3RVrflVNevxp4GNznrijp++qyruiDtvvXG34Xvj/vzan3X+dNxtWDbuHIbh/w/wd94ZjV+mAyn8ZnYYcBcQBh50zt0cRA4RST7ekXQ46BhJLeFP4DKzMHAv8HOgH3CSmfVLdA4RkVQVxKMX9wa+c8794JyrBJ4Cjg4gh4hISgqi8HcB5td6vcCftxEzO8vMJpnZpOLi4oSFExFp6YIo/Ju772mTy/TOuQecc0XOuaJ27dolIJaISGoIovAvALrWel0ILAogh4hISgqi8H8O9DKznmaWBpwIvBRADhGRlJTw2zmdc1Vmdh7wJt7tnA8756YnOoeISKoK5D5+59xrwGtBrFtEJNUlxYNYzKwYmLudH28LLG/EOImW7Pkh+bdB+YOX7NsQVP7uzrlN7o5JisK/I8xs0uaeQJMskj0/JP82KH/wkn0bmlv+IC7uiohIgFT4RURSTCoU/geCDrCDkj0/JP82KH/wkn0bmlX+Ft/GLyIiG0uFI34REalFhV9EJMW06MJvZoeZ2Uwz+87MLg86z7aY2cNmtszMptWa19rM3jaz2f44yIcfbZWZdTWz98xshplNN7ML/PlJsQ1mlmFmn5nZl37+6/35SZG/hpmFzWyKmb3iv062/HPM7Gszm2pmk/x5SbMNZlZgZs+a2bf+fwv7Nrf8LbbwJ+kDXx4BDqsz73LgXedcL+Bd/3VzVQX8wTm3KzAE+L3/N0+WbagADnbODQQGAYeZ2RCSJ3+NC4AZtV4nW36Ag5xzg2rd+55M23AX8IZzri8wEO//i+aV3znXIgdgX+DNWq+vAK4IOlc9cvcAptV6PRPo5E93AmYGnbEB2/IiMDIZtwHIAr4A9kmm/Hi93b4LHAy8koz/hoA5QNs685JiG4A84Ef8G2eaa/4We8RPPR/4kgQ6OOcWA/jj9gHnqRcz6wEMBj4libbBbyaZCiwD3nbOJVV+4E5gDBCvNS+Z8oP3fI63zGyymZ3lz0uWbdgJKAb+5Te3PWhm2TSz/C258NfrgS/S+MwsB3gOuNA5tzboPA3hnKt2zg3CO3Le28x2DzhSvZnZKGCZc25y0Fl20DDn3B54zbS/N7MDgg7UABFgD+DvzrnBwDqCbtbZjJZc+FvKA1+WmlknAH+8LOA8W2VmUbyi/4Rz7nl/dlJtA4BzbjUwHu+aS7LkHwYcZWZz8J5lfbCZPU7y5AfAObfIHy8DxuE9pztZtmEBsMA/UwR4Fm9H0Kzyt+TC31Ie+PIScLo/fTpeu3mzZGYGPATMcM7dXuutpNgGM2tnZgX+dCZwCPAtSZLfOXeFc67QOdcD79/7/5xzp5Ik+QHMLNvMcmumgZ8B00iSbXDOLQHmm1kff9YI4BuaW/6gL4Y08YWWw4FZwPfAVUHnqUfeJ4HFQAzvyOFMoA3exbrZ/rh10Dm3kn8/vOa0r4Cp/nB4smwDMACY4uefBlzrz0+K/HW2ZTgbLu4mTX68NvIv/WF6zX+3SbYNg4BJ/r+jF4BWzS2/umwQEUkxLbmpR0RENkOFX0Qkxajwi4ikGBV+EZEUo8IvIpJiVPhFADOr9nuDrBka7deWZtajdo+rIkGLBB1ApJkoc15XDSItno74RbbC7xv+r34//Z+Z2S7+/O5m9q6ZfeWPu/nzO5jZOL9P/y/NbKj/VWEz+6ffz/9b/i+DRQKhwi/iyazT1HNCrffWOuf2Bu7B6/0Sf/ox59wA4AlgrD9/LPC+8/r03wPv16cAvYB7nXO7AauBXzTp1ohshX65KwKYWalzLmcz8+fgPZzlB78DuiXOuTZmthyvf/WYP3+xc66tmRUDhc65ilrf0QOvi+de/uvLgKhz7sYEbJrIJnTEL7JtbgvTW1pmcypqTVej62sSIBV+kW07odZ4oj/9MV4PmACnAB/60+8C58BPD3XJS1RIkfrSUYeIJ9N/8laNN5xzNbd0ppvZp3gHSif580YDD5vZpXhPXPq1P/8C4AEzOxPvyP4cvB5XRZoNtfGLbIXfxl/knFsedBaRxqKmHhGRFKMjfhGRFKMjfhGRFKPCLyKSYlT4RURSjAq/iEiKUeEXEUkx/x/V8VmPp0pytAAAAABJRU5ErkJggg=="
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Let's try to visualize this on a time series plot"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 887,
   "source": [
    "# X_val_and_test"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 889,
   "source": [
    "# Y_val_and_test"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "source": [
    "actual = pd.DataFrame(Y_val_and_test, columns=[\"actual\"])\n",
    "actual.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "     actual\n",
       "0  5.736426\n",
       "1 -0.936343\n",
       "2  5.753206\n",
       "3 -0.575372\n",
       "4 -0.714855"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.736426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.936343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.753206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.575372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.714855</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 181
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "source": [
    "predicted = pd.DataFrame(model.predict(X_val_and_test), columns=[\"pred\"])\n",
    "predicted.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "       pred\n",
       "0  5.635177\n",
       "1 -0.937195\n",
       "2  5.373870\n",
       "3 -0.902014\n",
       "4 -0.860707"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.635177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.937195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.373870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.902014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.860707</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 182
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "source": [
    "actual_v_pred = actual.join(predicted, how = \"inner\")\n",
    "#actual_v_pred['pred_altered'] = actual_v_pred['pred'].apply(lambda x: (((x-1)/2)**2)*5)\n",
    "actual_v_pred.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "     actual      pred\n",
       "0  5.736426  5.635177\n",
       "1 -0.936343 -0.937195\n",
       "2  5.753206  5.373870\n",
       "3 -0.575372 -0.902014\n",
       "4 -0.714855 -0.860707"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.736426</td>\n",
       "      <td>5.635177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.936343</td>\n",
       "      <td>-0.937195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.753206</td>\n",
       "      <td>5.373870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.575372</td>\n",
       "      <td>-0.902014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.714855</td>\n",
       "      <td>-0.860707</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 183
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "source": [
    "plt.plot(actual_v_pred)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fad261d93a0>,\n",
       " <matplotlib.lines.Line2D at 0x7fad261d9490>]"
      ]
     },
     "metadata": {},
     "execution_count": 184
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "source": [
    "import numpy as np\n",
    "np.corrcoef(actual_v_pred['actual'], actual_v_pred['pred'])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "array([[1.       , 0.9734566],\n",
       "       [0.9734566, 1.       ]])"
      ]
     },
     "metadata": {},
     "execution_count": 185
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "source": [
    "from scipy.stats import pearsonr\n",
    "\n",
    "pearsonr(actual_v_pred['actual'], actual_v_pred['pred'])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(0.9734566033813612, 1.843898588292523e-30)"
      ]
     },
     "metadata": {},
     "execution_count": 186
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "This says that the two variables have a correlation of 0.97 and a p-value of basically 0. "
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "source": [
    "model.evaluate(X_train, Y_train), mean_squared_error(actual_v_pred['actual'], actual_v_pred['pred'])"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2202 - mean_squared_error: 0.2202\n"
     ]
    },
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "([0.22024847567081451, 0.22024847567081451], 0.4426634)"
      ]
     },
     "metadata": {},
     "execution_count": 187
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Let's try a 'regular' multiple linear regression model:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "source": [
    "from sklearn.linear_model import LinearRegression"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "source": [
    "model = LinearRegression()\n",
    "model.fit(X_train, Y_train)\n",
    "preds_linear = pd.DataFrame(model.predict(X_val_and_test), columns=[\"pred\"])\n",
    "#print(\"Loss: \" + str(loss(rounded_preds, test_labels)))"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "source": [
    "actual_v_pred_linear = actual.join(preds_linear, how = \"inner\")\n",
    "actual_v_pred_linear['diff'] = actual_v_pred_linear['actual'] - actual_v_pred_linear['pred']\n",
    "actual_v_pred_linear.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "     actual      pred      diff\n",
       "0  5.736426  5.698994  0.037432\n",
       "1 -0.936343 -0.812579 -0.123764\n",
       "2  5.753206  5.698608  0.054598\n",
       "3 -0.575372 -0.869005  0.293634\n",
       "4 -0.714855 -1.001359  0.286504"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>pred</th>\n",
       "      <th>diff</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.736426</td>\n",
       "      <td>5.698994</td>\n",
       "      <td>0.037432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.936343</td>\n",
       "      <td>-0.812579</td>\n",
       "      <td>-0.123764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.753206</td>\n",
       "      <td>5.698608</td>\n",
       "      <td>0.054598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.575372</td>\n",
       "      <td>-0.869005</td>\n",
       "      <td>0.293634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.714855</td>\n",
       "      <td>-1.001359</td>\n",
       "      <td>0.286504</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 190
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "model.score(X_train, Y_train), mean_squared_error(Y_val_and_test, preds_linear)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(0.9796442749598053, 0.33060735)"
      ]
     },
     "metadata": {},
     "execution_count": 191
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Here we see that the LinearRegression model produced a `score` of 0.979 and a MSE of 0.33."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "source": [
    "plt.scatter(actual_v_pred_linear['actual'], actual_v_pred_linear['pred'])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fad268337c0>"
      ]
     },
     "metadata": {},
     "execution_count": 192
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "source": [
    "plt.plot(actual_v_pred_linear['actual'])\n",
    "plt.plot(actual_v_pred_linear['pred'])"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fad23933040>]"
      ]
     },
     "metadata": {},
     "execution_count": 193
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "WOw, the predicted are so similar to the actual that you cannot discern between the two."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "source": [
    "print('Intercept: \\n', model.intercept_)\n",
    "print('Coefficients: \\n', model.coef_)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Intercept: \n",
      " 5.82712\n",
      "Coefficients: \n",
      " [ 0.3110296 -0.7779452 -1.8181679 -5.610072   1.9364651]\n"
     ]
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "If we assign the coefficients to their respective input features (`year`, `d18O`, `lat`, `long`, `depth`) it is weird to observe that latitude, longitude, and depth had the 'largest' impact on the output `temp` variable -- far more than `d18O`!\n",
    "\n",
    "Looking at this is one of the many benefits of using a LinReg Model, we are able to see what the model finds important (versus a NN where a lot is hidden).\n",
    "\n",
    "<!-- Perhaps as expected, the model was able to discern that the d18O was the most important factor in determining the predicted temperature. This is seen by the second coefficient in the model (relating to d18O) being the largest number. It appears as tho the other coefficients (relating to year, lat, long, and depth) are extraordinarily tiny comparatively. -->"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "source": [
    "paleo_df"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "    year      temp      d18O    lat    long  depth location\n",
       "0   1967 -1.301917 -0.132046  73.25  274.75      5       ab\n",
       "1   1968 -0.175512 -0.134891  73.25  274.75      5       ab\n",
       "2   1969 -0.575391 -0.137737  73.25  274.75      5       ab\n",
       "3   1970 -1.314653 -0.134845  73.25  274.75      5       ab\n",
       "4   1971 -1.344812 -0.131952  73.25  274.75      5       ab\n",
       "..   ...       ...       ...    ...     ...    ...      ...\n",
       "22  2004 -1.368374 -1.801891  67.56  295.98      5     qikB\n",
       "23  2005 -1.031412 -2.045915  67.56  295.98      5     qikB\n",
       "24  2006 -0.294162 -2.200304  67.56  295.98      5     qikB\n",
       "25  2007 -0.349986 -2.143647  67.56  295.98      5     qikB\n",
       "26  2008 -0.626585 -2.214013  67.56  295.98      5     qikB\n",
       "\n",
       "[233 rows x 7 columns]"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>temp</th>\n",
       "      <th>d18O</th>\n",
       "      <th>lat</th>\n",
       "      <th>long</th>\n",
       "      <th>depth</th>\n",
       "      <th>location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1967</td>\n",
       "      <td>-1.301917</td>\n",
       "      <td>-0.132046</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1968</td>\n",
       "      <td>-0.175512</td>\n",
       "      <td>-0.134891</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1969</td>\n",
       "      <td>-0.575391</td>\n",
       "      <td>-0.137737</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>-1.314653</td>\n",
       "      <td>-0.134845</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1971</td>\n",
       "      <td>-1.344812</td>\n",
       "      <td>-0.131952</td>\n",
       "      <td>73.25</td>\n",
       "      <td>274.75</td>\n",
       "      <td>5</td>\n",
       "      <td>ab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2004</td>\n",
       "      <td>-1.368374</td>\n",
       "      <td>-1.801891</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2005</td>\n",
       "      <td>-1.031412</td>\n",
       "      <td>-2.045915</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>2006</td>\n",
       "      <td>-0.294162</td>\n",
       "      <td>-2.200304</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>2007</td>\n",
       "      <td>-0.349986</td>\n",
       "      <td>-2.143647</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2008</td>\n",
       "      <td>-0.626585</td>\n",
       "      <td>-2.214013</td>\n",
       "      <td>67.56</td>\n",
       "      <td>295.98</td>\n",
       "      <td>5</td>\n",
       "      <td>qikB</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>233 rows × 7 columns</p>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 195
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Sequential Neural Network"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "Okay.. so linear regression seemed to do the trick for these organisms really damn well. Let's see if we can recreate this with a Sequential Neural Network (aka: a sequence of linear combinations as a result of matrix operations). We will be using the ReLU activation function ( a linear function )"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "source": [
    "network = Sequential()\n",
    "network.add(Dense(8, input_shape=(5,), activation='LeakyReLU'))\n",
    "network.add(Dense(6, activation='LeakyReLU'))\n",
    "network.add(Dense(6, activation='LeakyReLU'))\n",
    "network.add(Dense(4, activation='LeakyReLU'))\n",
    "network.add(Dense(1, activation='LeakyReLU'))\n",
    "\n",
    "network.compile('adam', loss='mse', metrics=['mse'])\n",
    "network.fit(X_train, Y_train, epochs=1000)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Epoch 1/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 11.0279 - mse: 11.0279\n",
      "Epoch 2/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 10.9302 - mse: 10.9302\n",
      "Epoch 3/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 10.8130 - mse: 10.8130\n",
      "Epoch 4/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 10.6833 - mse: 10.6833\n",
      "Epoch 5/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 10.5486 - mse: 10.5486\n",
      "Epoch 6/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 10.3828 - mse: 10.3828\n",
      "Epoch 7/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 10.1895 - mse: 10.1895\n",
      "Epoch 8/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 9.9919 - mse: 9.9919\n",
      "Epoch 9/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 9.7634 - mse: 9.7634\n",
      "Epoch 10/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 9.5325 - mse: 9.5325\n",
      "Epoch 11/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 9.2872 - mse: 9.2872\n",
      "Epoch 12/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 9.0149 - mse: 9.0149\n",
      "Epoch 13/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 8.6901 - mse: 8.6901\n",
      "Epoch 14/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 8.3161 - mse: 8.3161\n",
      "Epoch 15/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 7.8163 - mse: 7.8163\n",
      "Epoch 16/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 7.2887 - mse: 7.2887\n",
      "Epoch 17/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 6.7072 - mse: 6.7072\n",
      "Epoch 18/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 6.1210 - mse: 6.1210\n",
      "Epoch 19/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 5.4721 - mse: 5.4721\n",
      "Epoch 20/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 4.8121 - mse: 4.8121\n",
      "Epoch 21/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 4.2363 - mse: 4.2363\n",
      "Epoch 22/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 3.5744 - mse: 3.5744\n",
      "Epoch 23/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 2.9529 - mse: 2.9529\n",
      "Epoch 24/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 2.4364 - mse: 2.4364\n",
      "Epoch 25/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 1.9183 - mse: 1.9183\n",
      "Epoch 26/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 1.4946 - mse: 1.4946\n",
      "Epoch 27/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 1.1372 - mse: 1.1372\n",
      "Epoch 28/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.8636 - mse: 0.8636\n",
      "Epoch 29/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.7041 - mse: 0.7041\n",
      "Epoch 30/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.6053 - mse: 0.6053\n",
      "Epoch 31/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.5553 - mse: 0.5553\n",
      "Epoch 32/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.5366 - mse: 0.5366\n",
      "Epoch 33/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.5238 - mse: 0.5238\n",
      "Epoch 34/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.5154 - mse: 0.5154\n",
      "Epoch 35/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.5052 - mse: 0.5052\n",
      "Epoch 36/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4962 - mse: 0.4962\n",
      "Epoch 37/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4868 - mse: 0.4868\n",
      "Epoch 38/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4794 - mse: 0.4794\n",
      "Epoch 39/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4697 - mse: 0.4697\n",
      "Epoch 40/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4606 - mse: 0.4606\n",
      "Epoch 41/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4517 - mse: 0.4517\n",
      "Epoch 42/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4429 - mse: 0.4429\n",
      "Epoch 43/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4349 - mse: 0.4349\n",
      "Epoch 44/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4257 - mse: 0.4257\n",
      "Epoch 45/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4172 - mse: 0.4172\n",
      "Epoch 46/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.4078 - mse: 0.4078\n",
      "Epoch 47/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3995 - mse: 0.3995\n",
      "Epoch 48/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3901 - mse: 0.3901\n",
      "Epoch 49/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3811 - mse: 0.3811\n",
      "Epoch 50/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3747 - mse: 0.3747\n",
      "Epoch 51/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3633 - mse: 0.3633\n",
      "Epoch 52/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3544 - mse: 0.3544\n",
      "Epoch 53/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3453 - mse: 0.3453\n",
      "Epoch 54/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3366 - mse: 0.3366\n",
      "Epoch 55/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3274 - mse: 0.3274\n",
      "Epoch 56/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3195 - mse: 0.3195\n",
      "Epoch 57/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3114 - mse: 0.3114\n",
      "Epoch 58/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.3026 - mse: 0.3026\n",
      "Epoch 59/1000\n",
      "6/6 [==============================] - 0s 923us/step - loss: 0.2950 - mse: 0.2950\n",
      "Epoch 60/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2877 - mse: 0.2877\n",
      "Epoch 61/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2798 - mse: 0.2798\n",
      "Epoch 62/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2732 - mse: 0.2732\n",
      "Epoch 63/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2662 - mse: 0.2662\n",
      "Epoch 64/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.2588 - mse: 0.2588\n",
      "Epoch 65/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.2526 - mse: 0.2526\n",
      "Epoch 66/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2467 - mse: 0.2467\n",
      "Epoch 67/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2417 - mse: 0.2417\n",
      "Epoch 68/1000\n",
      "6/6 [==============================] - 0s 977us/step - loss: 0.2373 - mse: 0.2373\n",
      "Epoch 69/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2309 - mse: 0.2309\n",
      "Epoch 70/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2265 - mse: 0.2265\n",
      "Epoch 71/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2230 - mse: 0.2230\n",
      "Epoch 72/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2181 - mse: 0.2181\n",
      "Epoch 73/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2154 - mse: 0.2154\n",
      "Epoch 74/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2112 - mse: 0.2112\n",
      "Epoch 75/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2085 - mse: 0.2085\n",
      "Epoch 76/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.2055 - mse: 0.2055\n",
      "Epoch 77/1000\n",
      "6/6 [==============================] - 0s 987us/step - loss: 0.2032 - mse: 0.2032\n",
      "Epoch 78/1000\n",
      "6/6 [==============================] - 0s 901us/step - loss: 0.2015 - mse: 0.2015\n",
      "Epoch 79/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1993 - mse: 0.1993\n",
      "Epoch 80/1000\n",
      "6/6 [==============================] - 0s 990us/step - loss: 0.1976 - mse: 0.1976\n",
      "Epoch 81/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1961 - mse: 0.1961\n",
      "Epoch 82/1000\n",
      "6/6 [==============================] - 0s 978us/step - loss: 0.1962 - mse: 0.1962\n",
      "Epoch 83/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1940 - mse: 0.1940\n",
      "Epoch 84/1000\n",
      "6/6 [==============================] - 0s 983us/step - loss: 0.1931 - mse: 0.1931\n",
      "Epoch 85/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1920 - mse: 0.1920\n",
      "Epoch 86/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1908 - mse: 0.1908\n",
      "Epoch 87/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1903 - mse: 0.1903\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Epoch 88/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1896 - mse: 0.1896\n",
      "Epoch 89/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1895 - mse: 0.1895\n",
      "Epoch 90/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1886 - mse: 0.1886\n",
      "Epoch 91/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1898 - mse: 0.1898\n",
      "Epoch 92/1000\n",
      "6/6 [==============================] - 0s 974us/step - loss: 0.1882 - mse: 0.1882\n",
      "Epoch 93/1000\n",
      "6/6 [==============================] - 0s 1000us/step - loss: 0.1892 - mse: 0.1892\n",
      "Epoch 94/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1872 - mse: 0.1872\n",
      "Epoch 95/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1872 - mse: 0.1872\n",
      "Epoch 96/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1870 - mse: 0.1870\n",
      "Epoch 97/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1867 - mse: 0.1867\n",
      "Epoch 98/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1871 - mse: 0.1871\n",
      "Epoch 99/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1866 - mse: 0.1866\n",
      "Epoch 100/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1863 - mse: 0.1863\n",
      "Epoch 101/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1860 - mse: 0.1860\n",
      "Epoch 102/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1858 - mse: 0.1858\n",
      "Epoch 103/1000\n",
      "6/6 [==============================] - 0s 919us/step - loss: 0.1868 - mse: 0.1868\n",
      "Epoch 104/1000\n",
      "6/6 [==============================] - 0s 990us/step - loss: 0.1859 - mse: 0.1859\n",
      "Epoch 105/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1868 - mse: 0.1868\n",
      "Epoch 106/1000\n",
      "6/6 [==============================] - 0s 991us/step - loss: 0.1858 - mse: 0.1858\n",
      "Epoch 107/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1860 - mse: 0.1860\n",
      "Epoch 108/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1851 - mse: 0.1851\n",
      "Epoch 109/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1848 - mse: 0.1848\n",
      "Epoch 110/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1850 - mse: 0.1850\n",
      "Epoch 111/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1847 - mse: 0.1847\n",
      "Epoch 112/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1847 - mse: 0.1847\n",
      "Epoch 113/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1845 - mse: 0.1845\n",
      "Epoch 114/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1844 - mse: 0.1844\n",
      "Epoch 115/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1840 - mse: 0.1840\n",
      "Epoch 116/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1846 - mse: 0.1846\n",
      "Epoch 117/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1856 - mse: 0.1856\n",
      "Epoch 118/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1835 - mse: 0.1835\n",
      "Epoch 119/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1837 - mse: 0.1837\n",
      "Epoch 120/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1841 - mse: 0.1841\n",
      "Epoch 121/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1833 - mse: 0.1833\n",
      "Epoch 122/1000\n",
      "6/6 [==============================] - 0s 999us/step - loss: 0.1841 - mse: 0.1841\n",
      "Epoch 123/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1838 - mse: 0.1838\n",
      "Epoch 124/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1838 - mse: 0.1838\n",
      "Epoch 125/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1833 - mse: 0.1833\n",
      "Epoch 126/1000\n",
      "6/6 [==============================] - 0s 999us/step - loss: 0.1837 - mse: 0.1837\n",
      "Epoch 127/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1826 - mse: 0.1826\n",
      "Epoch 128/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1833 - mse: 0.1833\n",
      "Epoch 129/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1833 - mse: 0.1833\n",
      "Epoch 130/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1826 - mse: 0.1826\n",
      "Epoch 131/1000\n",
      "6/6 [==============================] - 0s 984us/step - loss: 0.1829 - mse: 0.1829\n",
      "Epoch 132/1000\n",
      "6/6 [==============================] - 0s 984us/step - loss: 0.1822 - mse: 0.1822\n",
      "Epoch 133/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1822 - mse: 0.1822\n",
      "Epoch 134/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1818 - mse: 0.1818\n",
      "Epoch 135/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1822 - mse: 0.1822\n",
      "Epoch 136/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1822 - mse: 0.1822\n",
      "Epoch 137/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1815 - mse: 0.1815\n",
      "Epoch 138/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1823 - mse: 0.1823\n",
      "Epoch 139/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1818 - mse: 0.1818\n",
      "Epoch 140/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1823 - mse: 0.1823\n",
      "Epoch 141/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1815 - mse: 0.1815\n",
      "Epoch 142/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1820 - mse: 0.1820\n",
      "Epoch 143/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1810 - mse: 0.1810\n",
      "Epoch 144/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1818 - mse: 0.1818\n",
      "Epoch 145/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.1818 - mse: 0.1818\n",
      "Epoch 146/1000\n",
      "6/6 [==============================] - 0s 8ms/step - loss: 0.1811 - mse: 0.1811\n",
      "Epoch 147/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.1810 - mse: 0.1810\n",
      "Epoch 148/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1815 - mse: 0.1815\n",
      "Epoch 149/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1826 - mse: 0.1826\n",
      "Epoch 150/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1817 - mse: 0.1817\n",
      "Epoch 151/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1810 - mse: 0.1810\n",
      "Epoch 152/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1809 - mse: 0.1809\n",
      "Epoch 153/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1802 - mse: 0.1802\n",
      "Epoch 154/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1808 - mse: 0.1808\n",
      "Epoch 155/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1803 - mse: 0.1803\n",
      "Epoch 156/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1807 - mse: 0.1807\n",
      "Epoch 157/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1804 - mse: 0.1804\n",
      "Epoch 158/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1803 - mse: 0.1803\n",
      "Epoch 159/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1813 - mse: 0.1813\n",
      "Epoch 160/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1817 - mse: 0.1817\n",
      "Epoch 161/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1803 - mse: 0.1803\n",
      "Epoch 162/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1801 - mse: 0.1801\n",
      "Epoch 163/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1798 - mse: 0.1798\n",
      "Epoch 164/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1794 - mse: 0.1794\n",
      "Epoch 165/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1800 - mse: 0.1800\n",
      "Epoch 166/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1796 - mse: 0.1796\n",
      "Epoch 167/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1796 - mse: 0.1796\n",
      "Epoch 168/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1796 - mse: 0.1796\n",
      "Epoch 169/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1793 - mse: 0.1793\n",
      "Epoch 170/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1793 - mse: 0.1793\n",
      "Epoch 171/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1797 - mse: 0.1797\n",
      "Epoch 172/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1801 - mse: 0.1801\n",
      "Epoch 173/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1812 - mse: 0.1812\n",
      "Epoch 174/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1794 - mse: 0.1794\n",
      "Epoch 175/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1815 - mse: 0.1815\n",
      "Epoch 176/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1796 - mse: 0.1796\n",
      "Epoch 177/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1793 - mse: 0.1793\n",
      "Epoch 178/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1797 - mse: 0.1797\n",
      "Epoch 179/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1792 - mse: 0.1792\n",
      "Epoch 180/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1789 - mse: 0.1789\n",
      "Epoch 181/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1794 - mse: 0.1794\n",
      "Epoch 182/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1787 - mse: 0.1787\n",
      "Epoch 183/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1798 - mse: 0.1798\n",
      "Epoch 184/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1797 - mse: 0.1797\n",
      "Epoch 185/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1804 - mse: 0.1804\n",
      "Epoch 186/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1777 - mse: 0.1777\n",
      "Epoch 187/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1799 - mse: 0.1799\n",
      "Epoch 188/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1796 - mse: 0.1796\n",
      "Epoch 189/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1789 - mse: 0.1789\n",
      "Epoch 190/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1792 - mse: 0.1792\n",
      "Epoch 191/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1783 - mse: 0.1783\n",
      "Epoch 192/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1791 - mse: 0.1791\n",
      "Epoch 193/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1777 - mse: 0.1777\n",
      "Epoch 194/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1783 - mse: 0.1783\n",
      "Epoch 195/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1785 - mse: 0.1785\n",
      "Epoch 196/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1779 - mse: 0.1779\n",
      "Epoch 197/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1783 - mse: 0.1783\n",
      "Epoch 198/1000\n",
      "6/6 [==============================] - 0s 953us/step - loss: 0.1788 - mse: 0.1788\n",
      "Epoch 199/1000\n",
      "6/6 [==============================] - 0s 948us/step - loss: 0.1777 - mse: 0.1777\n",
      "Epoch 200/1000\n",
      "6/6 [==============================] - 0s 923us/step - loss: 0.1774 - mse: 0.1774\n",
      "Epoch 201/1000\n",
      "6/6 [==============================] - 0s 997us/step - loss: 0.1774 - mse: 0.1774\n",
      "Epoch 202/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1787 - mse: 0.1787\n",
      "Epoch 203/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1770 - mse: 0.1770\n",
      "Epoch 204/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1783 - mse: 0.1783\n",
      "Epoch 205/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1788 - mse: 0.1788\n",
      "Epoch 206/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1778 - mse: 0.1778\n",
      "Epoch 207/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1774 - mse: 0.1774\n",
      "Epoch 208/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1771 - mse: 0.1771\n",
      "Epoch 209/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1768 - mse: 0.1768\n",
      "Epoch 210/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1785 - mse: 0.1785\n",
      "Epoch 211/1000\n",
      "6/6 [==============================] - 0s 909us/step - loss: 0.1768 - mse: 0.1768\n",
      "Epoch 212/1000\n",
      "6/6 [==============================] - 0s 960us/step - loss: 0.1780 - mse: 0.1780\n",
      "Epoch 213/1000\n",
      "6/6 [==============================] - 0s 931us/step - loss: 0.1771 - mse: 0.1771\n",
      "Epoch 214/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1766 - mse: 0.1766\n",
      "Epoch 215/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1784 - mse: 0.1784\n",
      "Epoch 216/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1771 - mse: 0.1771\n",
      "Epoch 217/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1765 - mse: 0.1765\n",
      "Epoch 218/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1761 - mse: 0.1761\n",
      "Epoch 219/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1759 - mse: 0.1759\n",
      "Epoch 220/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1762 - mse: 0.1762\n",
      "Epoch 221/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1760 - mse: 0.1760\n",
      "Epoch 222/1000\n",
      "6/6 [==============================] - 0s 974us/step - loss: 0.1761 - mse: 0.1761\n",
      "Epoch 223/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1781 - mse: 0.1781\n",
      "Epoch 224/1000\n",
      "6/6 [==============================] - 0s 973us/step - loss: 0.1759 - mse: 0.1759\n",
      "Epoch 225/1000\n",
      "6/6 [==============================] - 0s 995us/step - loss: 0.1766 - mse: 0.1766\n",
      "Epoch 226/1000\n",
      "6/6 [==============================] - 0s 976us/step - loss: 0.1756 - mse: 0.1756\n",
      "Epoch 227/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1760 - mse: 0.1760\n",
      "Epoch 228/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1758 - mse: 0.1758\n",
      "Epoch 229/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1760 - mse: 0.1760\n",
      "Epoch 230/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1756 - mse: 0.1756\n",
      "Epoch 231/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1771 - mse: 0.1771\n",
      "Epoch 232/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1759 - mse: 0.1759\n",
      "Epoch 233/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1758 - mse: 0.1758\n",
      "Epoch 234/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1760 - mse: 0.1760\n",
      "Epoch 235/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1754 - mse: 0.1754\n",
      "Epoch 236/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1752 - mse: 0.1752\n",
      "Epoch 237/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1752 - mse: 0.1752\n",
      "Epoch 238/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1768 - mse: 0.1768\n",
      "Epoch 239/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1763 - mse: 0.1763\n",
      "Epoch 240/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1760 - mse: 0.1760\n",
      "Epoch 241/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1756 - mse: 0.1756\n",
      "Epoch 242/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1755 - mse: 0.1755\n",
      "Epoch 243/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1775 - mse: 0.1775\n",
      "Epoch 244/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1763 - mse: 0.1763\n",
      "Epoch 245/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1758 - mse: 0.1758\n",
      "Epoch 246/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1765 - mse: 0.1765\n",
      "Epoch 247/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1758 - mse: 0.1758\n",
      "Epoch 248/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1761 - mse: 0.1761\n",
      "Epoch 249/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1756 - mse: 0.1756\n",
      "Epoch 250/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1747 - mse: 0.1747\n",
      "Epoch 251/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1747 - mse: 0.1747\n",
      "Epoch 252/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1748 - mse: 0.1748\n",
      "Epoch 253/1000\n",
      "6/6 [==============================] - 0s 907us/step - loss: 0.1748 - mse: 0.1748\n",
      "Epoch 254/1000\n",
      "6/6 [==============================] - 0s 912us/step - loss: 0.1769 - mse: 0.1769\n",
      "Epoch 255/1000\n",
      "6/6 [==============================] - 0s 982us/step - loss: 0.1769 - mse: 0.1769\n",
      "Epoch 256/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1757 - mse: 0.1757\n",
      "Epoch 257/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1764 - mse: 0.1764\n",
      "Epoch 258/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1755 - mse: 0.1755\n",
      "Epoch 259/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1742 - mse: 0.1742\n",
      "Epoch 260/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1749 - mse: 0.1749\n",
      "Epoch 261/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1751 - mse: 0.1751\n",
      "Epoch 262/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1751 - mse: 0.1751\n",
      "Epoch 263/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1750 - mse: 0.1750\n",
      "Epoch 264/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1750 - mse: 0.1750\n",
      "Epoch 265/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1747 - mse: 0.1747\n",
      "Epoch 266/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1763 - mse: 0.1763\n",
      "Epoch 267/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1751 - mse: 0.1751\n",
      "Epoch 268/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1745 - mse: 0.1745\n",
      "Epoch 269/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1751 - mse: 0.1751\n",
      "Epoch 270/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1750 - mse: 0.1750\n",
      "Epoch 271/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1746 - mse: 0.1746\n",
      "Epoch 272/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1739 - mse: 0.1739\n",
      "Epoch 273/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 274/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1740 - mse: 0.1740\n",
      "Epoch 275/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1746 - mse: 0.1746\n",
      "Epoch 276/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1740 - mse: 0.1740\n",
      "Epoch 277/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 278/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 279/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1736 - mse: 0.1736\n",
      "Epoch 280/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1748 - mse: 0.1748\n",
      "Epoch 281/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1734 - mse: 0.1734\n",
      "Epoch 282/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1739 - mse: 0.1739\n",
      "Epoch 283/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 284/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1739 - mse: 0.1739\n",
      "Epoch 285/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1736 - mse: 0.1736\n",
      "Epoch 286/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1740 - mse: 0.1740\n",
      "Epoch 287/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 288/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1736 - mse: 0.1736\n",
      "Epoch 289/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 290/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 291/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 292/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1734 - mse: 0.1734\n",
      "Epoch 293/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1751 - mse: 0.1751\n",
      "Epoch 294/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 295/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1734 - mse: 0.1734\n",
      "Epoch 296/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 297/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 298/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1746 - mse: 0.1746\n",
      "Epoch 299/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1743 - mse: 0.1743\n",
      "Epoch 300/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1733 - mse: 0.1733\n",
      "Epoch 301/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1744 - mse: 0.1744\n",
      "Epoch 302/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 303/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1743 - mse: 0.1743\n",
      "Epoch 304/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1736 - mse: 0.1736\n",
      "Epoch 305/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1739 - mse: 0.1739\n",
      "Epoch 306/1000\n",
      "6/6 [==============================] - 0s 940us/step - loss: 0.1734 - mse: 0.1734\n",
      "Epoch 307/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1754 - mse: 0.1754\n",
      "Epoch 308/1000\n",
      "6/6 [==============================] - 0s 973us/step - loss: 0.1763 - mse: 0.1763\n",
      "Epoch 309/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 310/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1728 - mse: 0.1728\n",
      "Epoch 311/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1763 - mse: 0.1763\n",
      "Epoch 312/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1750 - mse: 0.1750\n",
      "Epoch 313/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1729 - mse: 0.1729\n",
      "Epoch 314/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1727 - mse: 0.1727\n",
      "Epoch 315/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1744 - mse: 0.1744\n",
      "Epoch 316/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1746 - mse: 0.1746\n",
      "Epoch 317/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 318/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1724 - mse: 0.1724\n",
      "Epoch 319/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1727 - mse: 0.1727\n",
      "Epoch 320/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1745 - mse: 0.1745\n",
      "Epoch 321/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1739 - mse: 0.1739\n",
      "Epoch 322/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1737 - mse: 0.1737\n",
      "Epoch 323/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1727 - mse: 0.1727\n",
      "Epoch 324/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1733 - mse: 0.1733\n",
      "Epoch 325/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1730 - mse: 0.1730\n",
      "Epoch 326/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1727 - mse: 0.1727\n",
      "Epoch 327/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 328/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1734 - mse: 0.1734\n",
      "Epoch 329/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1728 - mse: 0.1728\n",
      "Epoch 330/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1731 - mse: 0.1731\n",
      "Epoch 331/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 332/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1724 - mse: 0.1724\n",
      "Epoch 333/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1724 - mse: 0.1724\n",
      "Epoch 334/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1730 - mse: 0.1730\n",
      "Epoch 335/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1724 - mse: 0.1724\n",
      "Epoch 336/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1732 - mse: 0.1732\n",
      "Epoch 337/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1730 - mse: 0.1730\n",
      "Epoch 338/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1733 - mse: 0.1733\n",
      "Epoch 339/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1723 - mse: 0.1723\n",
      "Epoch 340/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 341/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1739 - mse: 0.1739\n",
      "Epoch 342/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 343/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1732 - mse: 0.1732\n",
      "Epoch 344/1000\n",
      "6/6 [==============================] - 0s 965us/step - loss: 0.1719 - mse: 0.1719\n",
      "Epoch 345/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 346/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 347/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1741 - mse: 0.1741\n",
      "Epoch 348/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 349/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1709 - mse: 0.1709\n",
      "Epoch 350/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1742 - mse: 0.1742\n",
      "Epoch 351/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1729 - mse: 0.1729\n",
      "Epoch 352/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1720 - mse: 0.1720\n",
      "Epoch 353/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1736 - mse: 0.1736\n",
      "Epoch 354/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1721 - mse: 0.1721\n",
      "Epoch 355/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1729 - mse: 0.1729\n",
      "Epoch 356/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1738 - mse: 0.1738\n",
      "Epoch 357/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 358/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1719 - mse: 0.1719\n",
      "Epoch 359/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 360/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1730 - mse: 0.1730\n",
      "Epoch 361/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1718 - mse: 0.1718\n",
      "Epoch 362/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 363/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1729 - mse: 0.1729\n",
      "Epoch 364/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1724 - mse: 0.1724\n",
      "Epoch 365/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 366/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 367/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1731 - mse: 0.1731\n",
      "Epoch 368/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1719 - mse: 0.1719\n",
      "Epoch 369/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1730 - mse: 0.1730\n",
      "Epoch 370/1000\n",
      "6/6 [==============================] - 0s 975us/step - loss: 0.1736 - mse: 0.1736\n",
      "Epoch 371/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1755 - mse: 0.1755\n",
      "Epoch 372/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1719 - mse: 0.1719\n",
      "Epoch 373/1000\n",
      "6/6 [==============================] - 0s 993us/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 374/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1726 - mse: 0.1726\n",
      "Epoch 375/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1718 - mse: 0.1718\n",
      "Epoch 376/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1721 - mse: 0.1721\n",
      "Epoch 377/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1717 - mse: 0.1717\n",
      "Epoch 378/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1716 - mse: 0.1716\n",
      "Epoch 379/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1732 - mse: 0.1732\n",
      "Epoch 380/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 381/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1719 - mse: 0.1719\n",
      "Epoch 382/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1712 - mse: 0.1712\n",
      "Epoch 383/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1715 - mse: 0.1715\n",
      "Epoch 384/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1718 - mse: 0.1718\n",
      "Epoch 385/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1722 - mse: 0.1722\n",
      "Epoch 386/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1720 - mse: 0.1720\n",
      "Epoch 387/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1712 - mse: 0.1712\n",
      "Epoch 388/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1713 - mse: 0.1713\n",
      "Epoch 389/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1757 - mse: 0.1757\n",
      "Epoch 390/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1730 - mse: 0.1730\n",
      "Epoch 391/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1723 - mse: 0.1723\n",
      "Epoch 392/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1718 - mse: 0.1718\n",
      "Epoch 393/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1713 - mse: 0.1713\n",
      "Epoch 394/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1709 - mse: 0.1709\n",
      "Epoch 395/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1720 - mse: 0.1720\n",
      "Epoch 396/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1710 - mse: 0.1710\n",
      "Epoch 397/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1711 - mse: 0.1711\n",
      "Epoch 398/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1720 - mse: 0.1720\n",
      "Epoch 399/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1715 - mse: 0.1715\n",
      "Epoch 400/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1712 - mse: 0.1712\n",
      "Epoch 401/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1710 - mse: 0.1710\n",
      "Epoch 402/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1728 - mse: 0.1728\n",
      "Epoch 403/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 404/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 405/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1713 - mse: 0.1713\n",
      "Epoch 406/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1711 - mse: 0.1711\n",
      "Epoch 407/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1717 - mse: 0.1717\n",
      "Epoch 408/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1713 - mse: 0.1713\n",
      "Epoch 409/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1707 - mse: 0.1707\n",
      "Epoch 410/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1717 - mse: 0.1717\n",
      "Epoch 411/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 412/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1768 - mse: 0.1768\n",
      "Epoch 413/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 414/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1724 - mse: 0.1724\n",
      "Epoch 415/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1713 - mse: 0.1713\n",
      "Epoch 416/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1731 - mse: 0.1731\n",
      "Epoch 417/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1717 - mse: 0.1717\n",
      "Epoch 418/1000\n",
      "6/6 [==============================] - 0s 987us/step - loss: 0.1710 - mse: 0.1710\n",
      "Epoch 419/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1702 - mse: 0.1702\n",
      "Epoch 420/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1711 - mse: 0.1711\n",
      "Epoch 421/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 422/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1707 - mse: 0.1707\n",
      "Epoch 423/1000\n",
      "6/6 [==============================] - 0s 3ms/step - loss: 0.1728 - mse: 0.1728\n",
      "Epoch 424/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.1761 - mse: 0.1761\n",
      "Epoch 425/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1703 - mse: 0.1703\n",
      "Epoch 426/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1715 - mse: 0.1715\n",
      "Epoch 427/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1721 - mse: 0.1721\n",
      "Epoch 428/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1710 - mse: 0.1710\n",
      "Epoch 429/1000\n",
      "6/6 [==============================] - 0s 913us/step - loss: 0.1714 - mse: 0.1714\n",
      "Epoch 430/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1716 - mse: 0.1716\n",
      "Epoch 431/1000\n",
      "6/6 [==============================] - 0s 989us/step - loss: 0.1704 - mse: 0.1704\n",
      "Epoch 432/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1706 - mse: 0.1706\n",
      "Epoch 433/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1699 - mse: 0.1699\n",
      "Epoch 434/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1705 - mse: 0.1705\n",
      "Epoch 435/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1715 - mse: 0.1715\n",
      "Epoch 436/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1714 - mse: 0.1714\n",
      "Epoch 437/1000\n",
      "6/6 [==============================] - 0s 992us/step - loss: 0.1707 - mse: 0.1707\n",
      "Epoch 438/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1714 - mse: 0.1714\n",
      "Epoch 439/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1704 - mse: 0.1704\n",
      "Epoch 440/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 441/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1725 - mse: 0.1725\n",
      "Epoch 442/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1693 - mse: 0.1693\n",
      "Epoch 443/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1712 - mse: 0.1712\n",
      "Epoch 444/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1720 - mse: 0.1720\n",
      "Epoch 445/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1705 - mse: 0.1705\n",
      "Epoch 446/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1699 - mse: 0.1699\n",
      "Epoch 447/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1717 - mse: 0.1717\n",
      "Epoch 448/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1703 - mse: 0.1703\n",
      "Epoch 449/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1705 - mse: 0.1705\n",
      "Epoch 450/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1712 - mse: 0.1712\n",
      "Epoch 451/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 452/1000\n",
      "6/6 [==============================] - 0s 974us/step - loss: 0.1696 - mse: 0.1696\n",
      "Epoch 453/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1702 - mse: 0.1702\n",
      "Epoch 454/1000\n",
      "6/6 [==============================] - 0s 946us/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 455/1000\n",
      "6/6 [==============================] - 0s 979us/step - loss: 0.1699 - mse: 0.1699\n",
      "Epoch 456/1000\n",
      "6/6 [==============================] - 0s 998us/step - loss: 0.1694 - mse: 0.1694\n",
      "Epoch 457/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1700 - mse: 0.1700\n",
      "Epoch 458/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1698 - mse: 0.1698\n",
      "Epoch 459/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1704 - mse: 0.1704\n",
      "Epoch 460/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 461/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1695 - mse: 0.1695\n",
      "Epoch 462/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1701 - mse: 0.1701\n",
      "Epoch 463/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1700 - mse: 0.1700\n",
      "Epoch 464/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1695 - mse: 0.1695\n",
      "Epoch 465/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1693 - mse: 0.1693\n",
      "Epoch 466/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1694 - mse: 0.1694\n",
      "Epoch 467/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1698 - mse: 0.1698\n",
      "Epoch 468/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1702 - mse: 0.1702\n",
      "Epoch 469/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1699 - mse: 0.1699\n",
      "Epoch 470/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1701 - mse: 0.1701\n",
      "Epoch 471/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1706 - mse: 0.1706\n",
      "Epoch 472/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1698 - mse: 0.1698\n",
      "Epoch 473/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1695 - mse: 0.1695\n",
      "Epoch 474/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1690 - mse: 0.1690\n",
      "Epoch 475/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1699 - mse: 0.1699\n",
      "Epoch 476/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1694 - mse: 0.1694\n",
      "Epoch 477/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1698 - mse: 0.1698\n",
      "Epoch 478/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1693 - mse: 0.1693\n",
      "Epoch 479/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1717 - mse: 0.1717\n",
      "Epoch 480/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1692 - mse: 0.1692\n",
      "Epoch 481/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1692 - mse: 0.1692\n",
      "Epoch 482/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1694 - mse: 0.1694\n",
      "Epoch 483/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1692 - mse: 0.1692\n",
      "Epoch 484/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1705 - mse: 0.1705\n",
      "Epoch 485/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1718 - mse: 0.1718\n",
      "Epoch 486/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1698 - mse: 0.1698\n",
      "Epoch 487/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1723 - mse: 0.1723\n",
      "Epoch 488/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1690 - mse: 0.1690\n",
      "Epoch 489/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1696 - mse: 0.1696\n",
      "Epoch 490/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1686 - mse: 0.1686\n",
      "Epoch 491/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1682 - mse: 0.1682\n",
      "Epoch 492/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1740 - mse: 0.1740\n",
      "Epoch 493/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1679 - mse: 0.1679\n",
      "Epoch 494/1000\n",
      "6/6 [==============================] - 0s 998us/step - loss: 0.1715 - mse: 0.1715\n",
      "Epoch 495/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1710 - mse: 0.1710\n",
      "Epoch 496/1000\n",
      "6/6 [==============================] - 0s 972us/step - loss: 0.1709 - mse: 0.1709\n",
      "Epoch 497/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1681 - mse: 0.1681\n",
      "Epoch 498/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1713 - mse: 0.1713\n",
      "Epoch 499/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1714 - mse: 0.1714\n",
      "Epoch 500/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1689 - mse: 0.1689\n",
      "Epoch 501/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1691 - mse: 0.1691\n",
      "Epoch 502/1000\n",
      "6/6 [==============================] - 0s 963us/step - loss: 0.1696 - mse: 0.1696\n",
      "Epoch 503/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1681 - mse: 0.1681\n",
      "Epoch 504/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1684 - mse: 0.1684\n",
      "Epoch 505/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1691 - mse: 0.1691\n",
      "Epoch 506/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1690 - mse: 0.1690\n",
      "Epoch 507/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1692 - mse: 0.1692\n",
      "Epoch 508/1000\n",
      "6/6 [==============================] - 0s 996us/step - loss: 0.1683 - mse: 0.1683\n",
      "Epoch 509/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1688 - mse: 0.1688\n",
      "Epoch 510/1000\n",
      "6/6 [==============================] - 0s 947us/step - loss: 0.1681 - mse: 0.1681\n",
      "Epoch 511/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1678 - mse: 0.1678\n",
      "Epoch 512/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1680 - mse: 0.1680\n",
      "Epoch 513/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1683 - mse: 0.1683\n",
      "Epoch 514/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 515/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1685 - mse: 0.1685\n",
      "Epoch 516/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1690 - mse: 0.1690\n",
      "Epoch 517/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1689 - mse: 0.1689\n",
      "Epoch 518/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1674 - mse: 0.1674\n",
      "Epoch 519/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1684 - mse: 0.1684\n",
      "Epoch 520/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1679 - mse: 0.1679\n",
      "Epoch 521/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1688 - mse: 0.1688\n",
      "Epoch 522/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 523/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 524/1000\n",
      "6/6 [==============================] - 0s 963us/step - loss: 0.1706 - mse: 0.1706\n",
      "Epoch 525/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1675 - mse: 0.1675\n",
      "Epoch 526/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 527/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1694 - mse: 0.1694\n",
      "Epoch 528/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1678 - mse: 0.1678\n",
      "Epoch 529/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1678 - mse: 0.1678\n",
      "Epoch 530/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1690 - mse: 0.1690\n",
      "Epoch 531/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1688 - mse: 0.1688\n",
      "Epoch 532/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1701 - mse: 0.1701\n",
      "Epoch 533/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1679 - mse: 0.1679\n",
      "Epoch 534/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1695 - mse: 0.1695\n",
      "Epoch 535/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1691 - mse: 0.1691\n",
      "Epoch 536/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1686 - mse: 0.1686\n",
      "Epoch 537/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 538/1000\n",
      "6/6 [==============================] - 0s 896us/step - loss: 0.1685 - mse: 0.1685\n",
      "Epoch 539/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1672 - mse: 0.1672\n",
      "Epoch 540/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 541/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1669 - mse: 0.1669\n",
      "Epoch 542/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1673 - mse: 0.1673\n",
      "Epoch 543/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 544/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 545/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 546/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 547/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1679 - mse: 0.1679\n",
      "Epoch 548/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1677 - mse: 0.1677\n",
      "Epoch 549/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1669 - mse: 0.1669\n",
      "Epoch 550/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1675 - mse: 0.1675\n",
      "Epoch 551/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1669 - mse: 0.1669\n",
      "Epoch 552/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1683 - mse: 0.1683\n",
      "Epoch 553/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1671 - mse: 0.1671\n",
      "Epoch 554/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1683 - mse: 0.1683\n",
      "Epoch 555/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1670 - mse: 0.1670\n",
      "Epoch 556/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1687 - mse: 0.1687\n",
      "Epoch 557/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 558/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1672 - mse: 0.1672\n",
      "Epoch 559/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1670 - mse: 0.1670\n",
      "Epoch 560/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1679 - mse: 0.1679\n",
      "Epoch 561/1000\n",
      "6/6 [==============================] - 0s 918us/step - loss: 0.1697 - mse: 0.1697\n",
      "Epoch 562/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 563/1000\n",
      "6/6 [==============================] - 0s 900us/step - loss: 0.1665 - mse: 0.1665\n",
      "Epoch 564/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1685 - mse: 0.1685\n",
      "Epoch 565/1000\n",
      "6/6 [==============================] - 0s 859us/step - loss: 0.1678 - mse: 0.1678\n",
      "Epoch 566/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1677 - mse: 0.1677\n",
      "Epoch 567/1000\n",
      "6/6 [==============================] - 0s 970us/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 568/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1669 - mse: 0.1669\n",
      "Epoch 569/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 570/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1692 - mse: 0.1692\n",
      "Epoch 571/1000\n",
      "6/6 [==============================] - 0s 968us/step - loss: 0.1663 - mse: 0.1663\n",
      "Epoch 572/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1677 - mse: 0.1677\n",
      "Epoch 573/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1670 - mse: 0.1670\n",
      "Epoch 574/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1671 - mse: 0.1671\n",
      "Epoch 575/1000\n",
      "6/6 [==============================] - 0s 931us/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 576/1000\n",
      "6/6 [==============================] - 0s 935us/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 577/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1678 - mse: 0.1678\n",
      "Epoch 578/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 579/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1673 - mse: 0.1673\n",
      "Epoch 580/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1662 - mse: 0.1662\n",
      "Epoch 581/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1671 - mse: 0.1671\n",
      "Epoch 582/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1659 - mse: 0.1659\n",
      "Epoch 583/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1681 - mse: 0.1681\n",
      "Epoch 584/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1708 - mse: 0.1708\n",
      "Epoch 585/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1686 - mse: 0.1686\n",
      "Epoch 586/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 587/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1664 - mse: 0.1664\n",
      "Epoch 588/1000\n",
      "6/6 [==============================] - 0s 950us/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 589/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1660 - mse: 0.1660\n",
      "Epoch 590/1000\n",
      "6/6 [==============================] - 0s 901us/step - loss: 0.1661 - mse: 0.1661\n",
      "Epoch 591/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1670 - mse: 0.1670\n",
      "Epoch 592/1000\n",
      "6/6 [==============================] - 0s 954us/step - loss: 0.1675 - mse: 0.1675\n",
      "Epoch 593/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 594/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1662 - mse: 0.1662\n",
      "Epoch 595/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1665 - mse: 0.1665\n",
      "Epoch 596/1000\n",
      "6/6 [==============================] - 0s 961us/step - loss: 0.1675 - mse: 0.1675\n",
      "Epoch 597/1000\n",
      "6/6 [==============================] - 0s 992us/step - loss: 0.1676 - mse: 0.1676\n",
      "Epoch 598/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1660 - mse: 0.1660\n",
      "Epoch 599/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1665 - mse: 0.1665\n",
      "Epoch 600/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1658 - mse: 0.1658\n",
      "Epoch 601/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1695 - mse: 0.1695\n",
      "Epoch 602/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 603/1000\n",
      "6/6 [==============================] - 0s 931us/step - loss: 0.1658 - mse: 0.1658\n",
      "Epoch 604/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1670 - mse: 0.1670\n",
      "Epoch 605/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 606/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1697 - mse: 0.1697\n",
      "Epoch 607/1000\n",
      "6/6 [==============================] - 0s 977us/step - loss: 0.1655 - mse: 0.1655\n",
      "Epoch 608/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1691 - mse: 0.1691\n",
      "Epoch 609/1000\n",
      "6/6 [==============================] - 0s 997us/step - loss: 0.1662 - mse: 0.1662\n",
      "Epoch 610/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1687 - mse: 0.1687\n",
      "Epoch 611/1000\n",
      "6/6 [==============================] - 0s 989us/step - loss: 0.1682 - mse: 0.1682\n",
      "Epoch 612/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1675 - mse: 0.1675\n",
      "Epoch 613/1000\n",
      "6/6 [==============================] - 0s 938us/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 614/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1706 - mse: 0.1706\n",
      "Epoch 615/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1654 - mse: 0.1654\n",
      "Epoch 616/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 617/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1654 - mse: 0.1654\n",
      "Epoch 618/1000\n",
      "6/6 [==============================] - 0s 991us/step - loss: 0.1661 - mse: 0.1661\n",
      "Epoch 619/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1657 - mse: 0.1657\n",
      "Epoch 620/1000\n",
      "6/6 [==============================] - 0s 999us/step - loss: 0.1652 - mse: 0.1652\n",
      "Epoch 621/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 622/1000\n",
      "6/6 [==============================] - 0s 997us/step - loss: 0.1654 - mse: 0.1654\n",
      "Epoch 623/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1656 - mse: 0.1656\n",
      "Epoch 624/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1661 - mse: 0.1661\n",
      "Epoch 625/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1650 - mse: 0.1650\n",
      "Epoch 626/1000\n",
      "6/6 [==============================] - 0s 877us/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 627/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1650 - mse: 0.1650\n",
      "Epoch 628/1000\n",
      "6/6 [==============================] - 0s 909us/step - loss: 0.1667 - mse: 0.1667\n",
      "Epoch 629/1000\n",
      "6/6 [==============================] - 0s 929us/step - loss: 0.1668 - mse: 0.1668\n",
      "Epoch 630/1000\n",
      "6/6 [==============================] - 0s 930us/step - loss: 0.1666 - mse: 0.1666\n",
      "Epoch 631/1000\n",
      "6/6 [==============================] - 0s 913us/step - loss: 0.1653 - mse: 0.1653\n",
      "Epoch 632/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1654 - mse: 0.1654\n",
      "Epoch 633/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1650 - mse: 0.1650\n",
      "Epoch 634/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 635/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 636/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1654 - mse: 0.1654\n",
      "Epoch 637/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1665 - mse: 0.1665\n",
      "Epoch 638/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1675 - mse: 0.1675\n",
      "Epoch 639/1000\n",
      "6/6 [==============================] - 0s 975us/step - loss: 0.1655 - mse: 0.1655\n",
      "Epoch 640/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1647 - mse: 0.1647\n",
      "Epoch 641/1000\n",
      "6/6 [==============================] - 0s 948us/step - loss: 0.1651 - mse: 0.1651\n",
      "Epoch 642/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 643/1000\n",
      "6/6 [==============================] - 0s 989us/step - loss: 0.1666 - mse: 0.1666\n",
      "Epoch 644/1000\n",
      "6/6 [==============================] - 0s 993us/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 645/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1649 - mse: 0.1649\n",
      "Epoch 646/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1647 - mse: 0.1647\n",
      "Epoch 647/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1644 - mse: 0.1644\n",
      "Epoch 648/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1642 - mse: 0.1642\n",
      "Epoch 649/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1649 - mse: 0.1649\n",
      "Epoch 650/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 651/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1644 - mse: 0.1644\n",
      "Epoch 652/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1643 - mse: 0.1643\n",
      "Epoch 653/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1645 - mse: 0.1645\n",
      "Epoch 654/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1653 - mse: 0.1653\n",
      "Epoch 655/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1659 - mse: 0.1659\n",
      "Epoch 656/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1645 - mse: 0.1645\n",
      "Epoch 657/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1648 - mse: 0.1648\n",
      "Epoch 658/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1640 - mse: 0.1640\n",
      "Epoch 659/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1647 - mse: 0.1647\n",
      "Epoch 660/1000\n",
      "6/6 [==============================] - 0s 913us/step - loss: 0.1651 - mse: 0.1651\n",
      "Epoch 661/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1641 - mse: 0.1641\n",
      "Epoch 662/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1649 - mse: 0.1649\n",
      "Epoch 663/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1644 - mse: 0.1644\n",
      "Epoch 664/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1647 - mse: 0.1647\n",
      "Epoch 665/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1650 - mse: 0.1650\n",
      "Epoch 666/1000\n",
      "6/6 [==============================] - 0s 997us/step - loss: 0.1641 - mse: 0.1641\n",
      "Epoch 667/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1683 - mse: 0.1683\n",
      "Epoch 668/1000\n",
      "6/6 [==============================] - 0s 953us/step - loss: 0.1641 - mse: 0.1641\n",
      "Epoch 669/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1655 - mse: 0.1655\n",
      "Epoch 670/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1637 - mse: 0.1637\n",
      "Epoch 671/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1681 - mse: 0.1681\n",
      "Epoch 672/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1630 - mse: 0.1630\n",
      "Epoch 673/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1649 - mse: 0.1649\n",
      "Epoch 674/1000\n",
      "6/6 [==============================] - 0s 921us/step - loss: 0.1641 - mse: 0.1641\n",
      "Epoch 675/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1647 - mse: 0.1647\n",
      "Epoch 676/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1643 - mse: 0.1643\n",
      "Epoch 677/1000\n",
      "6/6 [==============================] - ETA: 0s - loss: 0.1737 - mse: 0.173 - 0s 1ms/step - loss: 0.1658 - mse: 0.1658\n",
      "Epoch 678/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1646 - mse: 0.1646\n",
      "Epoch 679/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1658 - mse: 0.1658\n",
      "Epoch 680/1000\n",
      "6/6 [==============================] - 0s 948us/step - loss: 0.1656 - mse: 0.1656\n",
      "Epoch 681/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1665 - mse: 0.1665\n",
      "Epoch 682/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1642 - mse: 0.1642\n",
      "Epoch 683/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1640 - mse: 0.1640\n",
      "Epoch 684/1000\n",
      "6/6 [==============================] - 0s 994us/step - loss: 0.1640 - mse: 0.1640\n",
      "Epoch 685/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1635 - mse: 0.1635\n",
      "Epoch 686/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1643 - mse: 0.1643\n",
      "Epoch 687/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1640 - mse: 0.1640\n",
      "Epoch 688/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1637 - mse: 0.1637\n",
      "Epoch 689/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1635 - mse: 0.1635\n",
      "Epoch 690/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 691/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1638 - mse: 0.1638\n",
      "Epoch 692/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1636 - mse: 0.1636\n",
      "Epoch 693/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1644 - mse: 0.1644\n",
      "Epoch 694/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1665 - mse: 0.1665\n",
      "Epoch 695/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1642 - mse: 0.1642\n",
      "Epoch 696/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1642 - mse: 0.1642\n",
      "Epoch 697/1000\n",
      "6/6 [==============================] - 0s 951us/step - loss: 0.1638 - mse: 0.1638\n",
      "Epoch 698/1000\n",
      "6/6 [==============================] - 0s 909us/step - loss: 0.1637 - mse: 0.1637\n",
      "Epoch 699/1000\n",
      "6/6 [==============================] - 0s 911us/step - loss: 0.1634 - mse: 0.1634\n",
      "Epoch 700/1000\n",
      "6/6 [==============================] - 0s 994us/step - loss: 0.1631 - mse: 0.1631\n",
      "Epoch 701/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1629 - mse: 0.1629\n",
      "Epoch 702/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1630 - mse: 0.1630\n",
      "Epoch 703/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1632 - mse: 0.1632\n",
      "Epoch 704/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1634 - mse: 0.1634\n",
      "Epoch 705/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 706/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1644 - mse: 0.1644\n",
      "Epoch 707/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1636 - mse: 0.1636\n",
      "Epoch 708/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1637 - mse: 0.1637\n",
      "Epoch 709/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1636 - mse: 0.1636\n",
      "Epoch 710/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1643 - mse: 0.1643\n",
      "Epoch 711/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1649 - mse: 0.1649\n",
      "Epoch 712/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1634 - mse: 0.1634\n",
      "Epoch 713/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1629 - mse: 0.1629\n",
      "Epoch 714/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 715/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1630 - mse: 0.1630\n",
      "Epoch 716/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1628 - mse: 0.1628\n",
      "Epoch 717/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1639 - mse: 0.1639\n",
      "Epoch 718/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1662 - mse: 0.1662\n",
      "Epoch 719/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1639 - mse: 0.1639\n",
      "Epoch 720/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 721/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1642 - mse: 0.1642\n",
      "Epoch 722/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1634 - mse: 0.1634\n",
      "Epoch 723/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1641 - mse: 0.1641\n",
      "Epoch 724/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1639 - mse: 0.1639\n",
      "Epoch 725/1000\n",
      "6/6 [==============================] - 0s 971us/step - loss: 0.1629 - mse: 0.1629\n",
      "Epoch 726/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1645 - mse: 0.1645\n",
      "Epoch 727/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1638 - mse: 0.1638\n",
      "Epoch 728/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1635 - mse: 0.1635\n",
      "Epoch 729/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1626 - mse: 0.1626\n",
      "Epoch 730/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1630 - mse: 0.1630\n",
      "Epoch 731/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1628 - mse: 0.1628\n",
      "Epoch 732/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1631 - mse: 0.1631\n",
      "Epoch 733/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 734/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 735/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1630 - mse: 0.1630\n",
      "Epoch 736/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1628 - mse: 0.1628\n",
      "Epoch 737/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1623 - mse: 0.1623\n",
      "Epoch 738/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 739/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1645 - mse: 0.1645\n",
      "Epoch 740/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1625 - mse: 0.1625\n",
      "Epoch 741/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1631 - mse: 0.1631\n",
      "Epoch 742/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1669 - mse: 0.1669\n",
      "Epoch 743/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1621 - mse: 0.1621\n",
      "Epoch 744/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1645 - mse: 0.1645\n",
      "Epoch 745/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1618 - mse: 0.1618\n",
      "Epoch 746/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1625 - mse: 0.1625\n",
      "Epoch 747/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1616 - mse: 0.1616\n",
      "Epoch 748/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1619 - mse: 0.1619\n",
      "Epoch 749/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1626 - mse: 0.1626\n",
      "Epoch 750/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1626 - mse: 0.1626\n",
      "Epoch 751/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 752/1000\n",
      "6/6 [==============================] - 0s 999us/step - loss: 0.1626 - mse: 0.1626\n",
      "Epoch 753/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1635 - mse: 0.1635\n",
      "Epoch 754/1000\n",
      "6/6 [==============================] - 0s 979us/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 755/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1635 - mse: 0.1635\n",
      "Epoch 756/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1638 - mse: 0.1638\n",
      "Epoch 757/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 758/1000\n",
      "6/6 [==============================] - 0s 998us/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 759/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1621 - mse: 0.1621\n",
      "Epoch 760/1000\n",
      "6/6 [==============================] - 0s 904us/step - loss: 0.1614 - mse: 0.1614\n",
      "Epoch 761/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1618 - mse: 0.1618\n",
      "Epoch 762/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1618 - mse: 0.1618\n",
      "Epoch 763/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1618 - mse: 0.1618\n",
      "Epoch 764/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1623 - mse: 0.1623\n",
      "Epoch 765/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1620 - mse: 0.1620\n",
      "Epoch 766/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1615 - mse: 0.1615\n",
      "Epoch 767/1000\n",
      "6/6 [==============================] - 0s 936us/step - loss: 0.1626 - mse: 0.1626\n",
      "Epoch 768/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 769/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 770/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1621 - mse: 0.1621\n",
      "Epoch 771/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1618 - mse: 0.1618\n",
      "Epoch 772/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1653 - mse: 0.1653\n",
      "Epoch 773/1000\n",
      "6/6 [==============================] - 0s 990us/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 774/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1649 - mse: 0.1649\n",
      "Epoch 775/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 980us/step - loss: 0.1629 - mse: 0.1629\n",
      "Epoch 776/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1633 - mse: 0.1633\n",
      "Epoch 777/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1635 - mse: 0.1635\n",
      "Epoch 778/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1605 - mse: 0.1605\n",
      "Epoch 779/1000\n",
      "6/6 [==============================] - 0s 945us/step - loss: 0.1629 - mse: 0.1629\n",
      "Epoch 780/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1624 - mse: 0.1624\n",
      "Epoch 781/1000\n",
      "6/6 [==============================] - 0s 920us/step - loss: 0.1619 - mse: 0.1619\n",
      "Epoch 782/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1618 - mse: 0.1618\n",
      "Epoch 783/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1612 - mse: 0.1612\n",
      "Epoch 784/1000\n",
      "6/6 [==============================] - 0s 995us/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 785/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1614 - mse: 0.1614\n",
      "Epoch 786/1000\n",
      "6/6 [==============================] - 0s 953us/step - loss: 0.1612 - mse: 0.1612\n",
      "Epoch 787/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1615 - mse: 0.1615\n",
      "Epoch 788/1000\n",
      "6/6 [==============================] - 0s 1000us/step - loss: 0.1657 - mse: 0.1657\n",
      "Epoch 789/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1612 - mse: 0.1612\n",
      "Epoch 790/1000\n",
      "6/6 [==============================] - 0s 974us/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 791/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1623 - mse: 0.1623\n",
      "Epoch 792/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1616 - mse: 0.1616\n",
      "Epoch 793/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1611 - mse: 0.1611\n",
      "Epoch 794/1000\n",
      "6/6 [==============================] - 0s 977us/step - loss: 0.1613 - mse: 0.1613\n",
      "Epoch 795/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1621 - mse: 0.1621\n",
      "Epoch 796/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1610 - mse: 0.1610\n",
      "Epoch 797/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1616 - mse: 0.1616\n",
      "Epoch 798/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1615 - mse: 0.1615\n",
      "Epoch 799/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1610 - mse: 0.1610\n",
      "Epoch 800/1000\n",
      "6/6 [==============================] - 0s 972us/step - loss: 0.1614 - mse: 0.1614\n",
      "Epoch 801/1000\n",
      "6/6 [==============================] - 0s 978us/step - loss: 0.1604 - mse: 0.1604\n",
      "Epoch 802/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1641 - mse: 0.1641\n",
      "Epoch 803/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1619 - mse: 0.1619\n",
      "Epoch 804/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1624 - mse: 0.1624\n",
      "Epoch 805/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1607 - mse: 0.1607\n",
      "Epoch 806/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 807/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1603 - mse: 0.1603\n",
      "Epoch 808/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1611 - mse: 0.1611\n",
      "Epoch 809/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1610 - mse: 0.1610\n",
      "Epoch 810/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 811/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1619 - mse: 0.1619\n",
      "Epoch 812/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1655 - mse: 0.1655\n",
      "Epoch 813/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1631 - mse: 0.1631\n",
      "Epoch 814/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1629 - mse: 0.1629\n",
      "Epoch 815/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1614 - mse: 0.1614\n",
      "Epoch 816/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1608 - mse: 0.1608\n",
      "Epoch 817/1000\n",
      "6/6 [==============================] - 0s 995us/step - loss: 0.1599 - mse: 0.1599\n",
      "Epoch 818/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1610 - mse: 0.1610\n",
      "Epoch 819/1000\n",
      "6/6 [==============================] - 0s 985us/step - loss: 0.1609 - mse: 0.1609\n",
      "Epoch 820/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1606 - mse: 0.1606\n",
      "Epoch 821/1000\n",
      "6/6 [==============================] - 0s 876us/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 822/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1620 - mse: 0.1620\n",
      "Epoch 823/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1613 - mse: 0.1613\n",
      "Epoch 824/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1601 - mse: 0.1601\n",
      "Epoch 825/1000\n",
      "6/6 [==============================] - 0s 933us/step - loss: 0.1606 - mse: 0.1606\n",
      "Epoch 826/1000\n",
      "6/6 [==============================] - 0s 960us/step - loss: 0.1599 - mse: 0.1599\n",
      "Epoch 827/1000\n",
      "6/6 [==============================] - 0s 992us/step - loss: 0.1617 - mse: 0.1617\n",
      "Epoch 828/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1602 - mse: 0.1602\n",
      "Epoch 829/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1606 - mse: 0.1606\n",
      "Epoch 830/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1619 - mse: 0.1619\n",
      "Epoch 831/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1609 - mse: 0.1609\n",
      "Epoch 832/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1597 - mse: 0.1597\n",
      "Epoch 833/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1614 - mse: 0.1614\n",
      "Epoch 834/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1601 - mse: 0.1601\n",
      "Epoch 835/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1600 - mse: 0.1600\n",
      "Epoch 836/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1612 - mse: 0.1612\n",
      "Epoch 837/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1600 - mse: 0.1600\n",
      "Epoch 838/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1606 - mse: 0.1606\n",
      "Epoch 839/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1599 - mse: 0.1599\n",
      "Epoch 840/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1591 - mse: 0.1591\n",
      "Epoch 841/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1612 - mse: 0.1612\n",
      "Epoch 842/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1611 - mse: 0.1611\n",
      "Epoch 843/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1610 - mse: 0.1610\n",
      "Epoch 844/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1598 - mse: 0.1598\n",
      "Epoch 845/1000\n",
      "6/6 [==============================] - 0s 961us/step - loss: 0.1600 - mse: 0.1600\n",
      "Epoch 846/1000\n",
      "6/6 [==============================] - 0s 936us/step - loss: 0.1621 - mse: 0.1621\n",
      "Epoch 847/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1602 - mse: 0.1602\n",
      "Epoch 848/1000\n",
      "6/6 [==============================] - 0s 989us/step - loss: 0.1605 - mse: 0.1605\n",
      "Epoch 849/1000\n",
      "6/6 [==============================] - 0s 965us/step - loss: 0.1597 - mse: 0.1597\n",
      "Epoch 850/1000\n",
      "6/6 [==============================] - 0s 955us/step - loss: 0.1596 - mse: 0.1596\n",
      "Epoch 851/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1602 - mse: 0.1602\n",
      "Epoch 852/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1590 - mse: 0.1590\n",
      "Epoch 853/1000\n",
      "6/6 [==============================] - 0s 2ms/step - loss: 0.1595 - mse: 0.1595\n",
      "Epoch 854/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1601 - mse: 0.1601\n",
      "Epoch 855/1000\n",
      "6/6 [==============================] - 0s 3ms/step - loss: 0.1591 - mse: 0.1591\n",
      "Epoch 856/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1590 - mse: 0.1590\n",
      "Epoch 857/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1593 - mse: 0.1593\n",
      "Epoch 858/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1588 - mse: 0.1588\n",
      "Epoch 859/1000\n",
      "6/6 [==============================] - 0s 920us/step - loss: 0.1600 - mse: 0.1600\n",
      "Epoch 860/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 968us/step - loss: 0.1622 - mse: 0.1622\n",
      "Epoch 861/1000\n",
      "6/6 [==============================] - 0s 991us/step - loss: 0.1591 - mse: 0.1591\n",
      "Epoch 862/1000\n",
      "6/6 [==============================] - 0s 942us/step - loss: 0.1597 - mse: 0.1597\n",
      "Epoch 863/1000\n",
      "6/6 [==============================] - 0s 977us/step - loss: 0.1593 - mse: 0.1593\n",
      "Epoch 864/1000\n",
      "6/6 [==============================] - 0s 909us/step - loss: 0.1590 - mse: 0.1590\n",
      "Epoch 865/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1602 - mse: 0.1602\n",
      "Epoch 866/1000\n",
      "6/6 [==============================] - 0s 879us/step - loss: 0.1598 - mse: 0.1598\n",
      "Epoch 867/1000\n",
      "6/6 [==============================] - 0s 959us/step - loss: 0.1592 - mse: 0.1592\n",
      "Epoch 868/1000\n",
      "6/6 [==============================] - 0s 911us/step - loss: 0.1590 - mse: 0.1590\n",
      "Epoch 869/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1596 - mse: 0.1596\n",
      "Epoch 870/1000\n",
      "6/6 [==============================] - 0s 938us/step - loss: 0.1608 - mse: 0.1608\n",
      "Epoch 871/1000\n",
      "6/6 [==============================] - 0s 886us/step - loss: 0.1604 - mse: 0.1604\n",
      "Epoch 872/1000\n",
      "6/6 [==============================] - 0s 848us/step - loss: 0.1592 - mse: 0.1592\n",
      "Epoch 873/1000\n",
      "6/6 [==============================] - 0s 902us/step - loss: 0.1602 - mse: 0.1602\n",
      "Epoch 874/1000\n",
      "6/6 [==============================] - 0s 875us/step - loss: 0.1581 - mse: 0.1581\n",
      "Epoch 875/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1590 - mse: 0.1590\n",
      "Epoch 876/1000\n",
      "6/6 [==============================] - 0s 923us/step - loss: 0.1607 - mse: 0.1607\n",
      "Epoch 877/1000\n",
      "6/6 [==============================] - 0s 879us/step - loss: 0.1601 - mse: 0.1601\n",
      "Epoch 878/1000\n",
      "6/6 [==============================] - 0s 878us/step - loss: 0.1587 - mse: 0.1587\n",
      "Epoch 879/1000\n",
      "6/6 [==============================] - 0s 882us/step - loss: 0.1592 - mse: 0.1592\n",
      "Epoch 880/1000\n",
      "6/6 [==============================] - 0s 929us/step - loss: 0.1596 - mse: 0.1596\n",
      "Epoch 881/1000\n",
      "6/6 [==============================] - 0s 895us/step - loss: 0.1584 - mse: 0.1584\n",
      "Epoch 882/1000\n",
      "6/6 [==============================] - 0s 989us/step - loss: 0.1581 - mse: 0.1581\n",
      "Epoch 883/1000\n",
      "6/6 [==============================] - 0s 881us/step - loss: 0.1585 - mse: 0.1585\n",
      "Epoch 884/1000\n",
      "6/6 [==============================] - 0s 956us/step - loss: 0.1588 - mse: 0.1588\n",
      "Epoch 885/1000\n",
      "6/6 [==============================] - 0s 888us/step - loss: 0.1584 - mse: 0.1584\n",
      "Epoch 886/1000\n",
      "6/6 [==============================] - 0s 881us/step - loss: 0.1582 - mse: 0.1582\n",
      "Epoch 887/1000\n",
      "6/6 [==============================] - 0s 972us/step - loss: 0.1590 - mse: 0.1590\n",
      "Epoch 888/1000\n",
      "6/6 [==============================] - 0s 916us/step - loss: 0.1627 - mse: 0.1627\n",
      "Epoch 889/1000\n",
      "6/6 [==============================] - 0s 977us/step - loss: 0.1580 - mse: 0.1580\n",
      "Epoch 890/1000\n",
      "6/6 [==============================] - 0s 962us/step - loss: 0.1592 - mse: 0.1592\n",
      "Epoch 891/1000\n",
      "6/6 [==============================] - 0s 983us/step - loss: 0.1601 - mse: 0.1601\n",
      "Epoch 892/1000\n",
      "6/6 [==============================] - 0s 908us/step - loss: 0.1583 - mse: 0.1583\n",
      "Epoch 893/1000\n",
      "6/6 [==============================] - 0s 937us/step - loss: 0.1584 - mse: 0.1584\n",
      "Epoch 894/1000\n",
      "6/6 [==============================] - 0s 946us/step - loss: 0.1578 - mse: 0.1578\n",
      "Epoch 895/1000\n",
      "6/6 [==============================] - 0s 906us/step - loss: 0.1589 - mse: 0.1589\n",
      "Epoch 896/1000\n",
      "6/6 [==============================] - 0s 895us/step - loss: 0.1595 - mse: 0.1595\n",
      "Epoch 897/1000\n",
      "6/6 [==============================] - 0s 908us/step - loss: 0.1598 - mse: 0.1598\n",
      "Epoch 898/1000\n",
      "6/6 [==============================] - 0s 933us/step - loss: 0.1588 - mse: 0.1588\n",
      "Epoch 899/1000\n",
      "6/6 [==============================] - 0s 923us/step - loss: 0.1576 - mse: 0.1576\n",
      "Epoch 900/1000\n",
      "6/6 [==============================] - 0s 887us/step - loss: 0.1586 - mse: 0.1586\n",
      "Epoch 901/1000\n",
      "6/6 [==============================] - 0s 991us/step - loss: 0.1584 - mse: 0.1584\n",
      "Epoch 902/1000\n",
      "6/6 [==============================] - 0s 930us/step - loss: 0.1578 - mse: 0.1578\n",
      "Epoch 903/1000\n",
      "6/6 [==============================] - 0s 850us/step - loss: 0.1579 - mse: 0.1579\n",
      "Epoch 904/1000\n",
      "6/6 [==============================] - 0s 917us/step - loss: 0.1582 - mse: 0.1582\n",
      "Epoch 905/1000\n",
      "6/6 [==============================] - 0s 870us/step - loss: 0.1578 - mse: 0.1578\n",
      "Epoch 906/1000\n",
      "6/6 [==============================] - 0s 938us/step - loss: 0.1572 - mse: 0.1572\n",
      "Epoch 907/1000\n",
      "6/6 [==============================] - 0s 924us/step - loss: 0.1577 - mse: 0.1577\n",
      "Epoch 908/1000\n",
      "6/6 [==============================] - 0s 915us/step - loss: 0.1574 - mse: 0.1574\n",
      "Epoch 909/1000\n",
      "6/6 [==============================] - 0s 904us/step - loss: 0.1577 - mse: 0.1577\n",
      "Epoch 910/1000\n",
      "6/6 [==============================] - 0s 882us/step - loss: 0.1577 - mse: 0.1577\n",
      "Epoch 911/1000\n",
      "6/6 [==============================] - 0s 960us/step - loss: 0.1577 - mse: 0.1577\n",
      "Epoch 912/1000\n",
      "6/6 [==============================] - 0s 896us/step - loss: 0.1589 - mse: 0.1589\n",
      "Epoch 913/1000\n",
      "6/6 [==============================] - 0s 956us/step - loss: 0.1580 - mse: 0.1580\n",
      "Epoch 914/1000\n",
      "6/6 [==============================] - 0s 916us/step - loss: 0.1608 - mse: 0.1608\n",
      "Epoch 915/1000\n",
      "6/6 [==============================] - 0s 904us/step - loss: 0.1562 - mse: 0.1562\n",
      "Epoch 916/1000\n",
      "6/6 [==============================] - 0s 953us/step - loss: 0.1575 - mse: 0.1575\n",
      "Epoch 917/1000\n",
      "6/6 [==============================] - 0s 995us/step - loss: 0.1571 - mse: 0.1571\n",
      "Epoch 918/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1585 - mse: 0.1585\n",
      "Epoch 919/1000\n",
      "6/6 [==============================] - 0s 998us/step - loss: 0.1596 - mse: 0.1596\n",
      "Epoch 920/1000\n",
      "6/6 [==============================] - 0s 954us/step - loss: 0.1576 - mse: 0.1576\n",
      "Epoch 921/1000\n",
      "6/6 [==============================] - 0s 869us/step - loss: 0.1581 - mse: 0.1581\n",
      "Epoch 922/1000\n",
      "6/6 [==============================] - 0s 938us/step - loss: 0.1569 - mse: 0.1569\n",
      "Epoch 923/1000\n",
      "6/6 [==============================] - 0s 982us/step - loss: 0.1578 - mse: 0.1578\n",
      "Epoch 924/1000\n",
      "6/6 [==============================] - 0s 949us/step - loss: 0.1574 - mse: 0.1574\n",
      "Epoch 925/1000\n",
      "6/6 [==============================] - 0s 950us/step - loss: 0.1572 - mse: 0.1572\n",
      "Epoch 926/1000\n",
      "6/6 [==============================] - 0s 932us/step - loss: 0.1586 - mse: 0.1586\n",
      "Epoch 927/1000\n",
      "6/6 [==============================] - 0s 909us/step - loss: 0.1582 - mse: 0.1582\n",
      "Epoch 928/1000\n",
      "6/6 [==============================] - 0s 900us/step - loss: 0.1595 - mse: 0.1595\n",
      "Epoch 929/1000\n",
      "6/6 [==============================] - 0s 918us/step - loss: 0.1566 - mse: 0.1566\n",
      "Epoch 930/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1563 - mse: 0.1563\n",
      "Epoch 931/1000\n",
      "6/6 [==============================] - 0s 918us/step - loss: 0.1569 - mse: 0.1569\n",
      "Epoch 932/1000\n",
      "6/6 [==============================] - 0s 896us/step - loss: 0.1575 - mse: 0.1575\n",
      "Epoch 933/1000\n",
      "6/6 [==============================] - 0s 911us/step - loss: 0.1596 - mse: 0.1596\n",
      "Epoch 934/1000\n",
      "6/6 [==============================] - 0s 861us/step - loss: 0.1578 - mse: 0.1578\n",
      "Epoch 935/1000\n",
      "6/6 [==============================] - 0s 989us/step - loss: 0.1583 - mse: 0.1583\n",
      "Epoch 936/1000\n",
      "6/6 [==============================] - 0s 890us/step - loss: 0.1573 - mse: 0.1573\n",
      "Epoch 937/1000\n",
      "6/6 [==============================] - 0s 928us/step - loss: 0.1576 - mse: 0.1576\n",
      "Epoch 938/1000\n",
      "6/6 [==============================] - 0s 905us/step - loss: 0.1572 - mse: 0.1572\n",
      "Epoch 939/1000\n",
      "6/6 [==============================] - 0s 899us/step - loss: 0.1566 - mse: 0.1566\n",
      "Epoch 940/1000\n",
      "6/6 [==============================] - 0s 971us/step - loss: 0.1575 - mse: 0.1575\n",
      "Epoch 941/1000\n",
      "6/6 [==============================] - 0s 905us/step - loss: 0.1579 - mse: 0.1579\n",
      "Epoch 942/1000\n",
      "6/6 [==============================] - 0s 942us/step - loss: 0.1562 - mse: 0.1562\n",
      "Epoch 943/1000\n",
      "6/6 [==============================] - 0s 893us/step - loss: 0.1577 - mse: 0.1577\n",
      "Epoch 944/1000\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "6/6 [==============================] - 0s 891us/step - loss: 0.1559 - mse: 0.1559\n",
      "Epoch 945/1000\n",
      "6/6 [==============================] - 0s 906us/step - loss: 0.1584 - mse: 0.1584\n",
      "Epoch 946/1000\n",
      "6/6 [==============================] - 0s 880us/step - loss: 0.1565 - mse: 0.1565\n",
      "Epoch 947/1000\n",
      "6/6 [==============================] - 0s 924us/step - loss: 0.1568 - mse: 0.1568\n",
      "Epoch 948/1000\n",
      "6/6 [==============================] - 0s 890us/step - loss: 0.1565 - mse: 0.1565\n",
      "Epoch 949/1000\n",
      "6/6 [==============================] - 0s 924us/step - loss: 0.1611 - mse: 0.1611\n",
      "Epoch 950/1000\n",
      "6/6 [==============================] - 0s 925us/step - loss: 0.1547 - mse: 0.1547\n",
      "Epoch 951/1000\n",
      "6/6 [==============================] - 0s 976us/step - loss: 0.1614 - mse: 0.1614\n",
      "Epoch 952/1000\n",
      "6/6 [==============================] - 0s 963us/step - loss: 0.1566 - mse: 0.1566\n",
      "Epoch 953/1000\n",
      "6/6 [==============================] - 0s 911us/step - loss: 0.1598 - mse: 0.1598\n",
      "Epoch 954/1000\n",
      "6/6 [==============================] - 0s 961us/step - loss: 0.1576 - mse: 0.1576\n",
      "Epoch 955/1000\n",
      "6/6 [==============================] - 0s 948us/step - loss: 0.1588 - mse: 0.1588\n",
      "Epoch 956/1000\n",
      "6/6 [==============================] - 0s 908us/step - loss: 0.1551 - mse: 0.1551\n",
      "Epoch 957/1000\n",
      "6/6 [==============================] - 0s 889us/step - loss: 0.1586 - mse: 0.1586\n",
      "Epoch 958/1000\n",
      "6/6 [==============================] - 0s 851us/step - loss: 0.1563 - mse: 0.1563\n",
      "Epoch 959/1000\n",
      "6/6 [==============================] - 0s 973us/step - loss: 0.1575 - mse: 0.1575\n",
      "Epoch 960/1000\n",
      "6/6 [==============================] - 0s 952us/step - loss: 0.1584 - mse: 0.1584\n",
      "Epoch 961/1000\n",
      "6/6 [==============================] - 0s 860us/step - loss: 0.1583 - mse: 0.1583\n",
      "Epoch 962/1000\n",
      "6/6 [==============================] - 0s 859us/step - loss: 0.1615 - mse: 0.1615\n",
      "Epoch 963/1000\n",
      "6/6 [==============================] - 0s 916us/step - loss: 0.1556 - mse: 0.1556\n",
      "Epoch 964/1000\n",
      "6/6 [==============================] - 0s 952us/step - loss: 0.1580 - mse: 0.1580\n",
      "Epoch 965/1000\n",
      "6/6 [==============================] - 0s 894us/step - loss: 0.1566 - mse: 0.1566\n",
      "Epoch 966/1000\n",
      "6/6 [==============================] - 0s 981us/step - loss: 0.1576 - mse: 0.1576\n",
      "Epoch 967/1000\n",
      "6/6 [==============================] - 0s 921us/step - loss: 0.1571 - mse: 0.1571\n",
      "Epoch 968/1000\n",
      "6/6 [==============================] - 0s 882us/step - loss: 0.1565 - mse: 0.1565\n",
      "Epoch 969/1000\n",
      "6/6 [==============================] - 0s 903us/step - loss: 0.1561 - mse: 0.1561\n",
      "Epoch 970/1000\n",
      "6/6 [==============================] - 0s 931us/step - loss: 0.1552 - mse: 0.1552\n",
      "Epoch 971/1000\n",
      "6/6 [==============================] - 0s 912us/step - loss: 0.1576 - mse: 0.1576\n",
      "Epoch 972/1000\n",
      "6/6 [==============================] - 0s 877us/step - loss: 0.1562 - mse: 0.1562\n",
      "Epoch 973/1000\n",
      "6/6 [==============================] - 0s 926us/step - loss: 0.1556 - mse: 0.1556\n",
      "Epoch 974/1000\n",
      "6/6 [==============================] - 0s 894us/step - loss: 0.1554 - mse: 0.1554\n",
      "Epoch 975/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1555 - mse: 0.1555\n",
      "Epoch 976/1000\n",
      "6/6 [==============================] - 0s 964us/step - loss: 0.1565 - mse: 0.1565\n",
      "Epoch 977/1000\n",
      "6/6 [==============================] - 0s 901us/step - loss: 0.1571 - mse: 0.1571\n",
      "Epoch 978/1000\n",
      "6/6 [==============================] - 0s 957us/step - loss: 0.1557 - mse: 0.1557\n",
      "Epoch 979/1000\n",
      "6/6 [==============================] - 0s 888us/step - loss: 0.1551 - mse: 0.1551\n",
      "Epoch 980/1000\n",
      "6/6 [==============================] - 0s 903us/step - loss: 0.1556 - mse: 0.1556\n",
      "Epoch 981/1000\n",
      "6/6 [==============================] - 0s 924us/step - loss: 0.1553 - mse: 0.1553\n",
      "Epoch 982/1000\n",
      "6/6 [==============================] - 0s 883us/step - loss: 0.1571 - mse: 0.1571\n",
      "Epoch 983/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1554 - mse: 0.1554\n",
      "Epoch 984/1000\n",
      "6/6 [==============================] - 0s 906us/step - loss: 0.1578 - mse: 0.1578\n",
      "Epoch 985/1000\n",
      "6/6 [==============================] - 0s 930us/step - loss: 0.1541 - mse: 0.1541\n",
      "Epoch 986/1000\n",
      "6/6 [==============================] - 0s 944us/step - loss: 0.1573 - mse: 0.1573\n",
      "Epoch 987/1000\n",
      "6/6 [==============================] - 0s 932us/step - loss: 0.1553 - mse: 0.1553\n",
      "Epoch 988/1000\n",
      "6/6 [==============================] - 0s 939us/step - loss: 0.1581 - mse: 0.1581\n",
      "Epoch 989/1000\n",
      "6/6 [==============================] - 0s 906us/step - loss: 0.1548 - mse: 0.1548\n",
      "Epoch 990/1000\n",
      "6/6 [==============================] - 0s 883us/step - loss: 0.1550 - mse: 0.1550\n",
      "Epoch 991/1000\n",
      "6/6 [==============================] - 0s 894us/step - loss: 0.1547 - mse: 0.1547\n",
      "Epoch 992/1000\n",
      "6/6 [==============================] - 0s 912us/step - loss: 0.1547 - mse: 0.1547\n",
      "Epoch 993/1000\n",
      "6/6 [==============================] - 0s 982us/step - loss: 0.1557 - mse: 0.1557\n",
      "Epoch 994/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1572 - mse: 0.1572\n",
      "Epoch 995/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1554 - mse: 0.1554\n",
      "Epoch 996/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1549 - mse: 0.1549\n",
      "Epoch 997/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1550 - mse: 0.1550\n",
      "Epoch 998/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1542 - mse: 0.1542\n",
      "Epoch 999/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1551 - mse: 0.1551\n",
      "Epoch 1000/1000\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.1555 - mse: 0.1555\n"
     ]
    },
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fad23357a60>"
      ]
     },
     "metadata": {},
     "execution_count": 196
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "source": [
    "predicted = pd.DataFrame(network.predict(X_val_and_test), columns=[\"pred\"])\n",
    "predicted.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "       pred\n",
       "0  5.697541\n",
       "1 -0.931081\n",
       "2  5.709077\n",
       "3 -0.929255\n",
       "4 -0.951365"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.697541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.931081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.709077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.929255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.951365</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 197
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "source": [
    "actual_v_pred = actual.join(predicted, how = \"inner\")\n",
    "actual_v_pred.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "     actual      pred\n",
       "0  5.736426  5.697541\n",
       "1 -0.936343 -0.931081\n",
       "2  5.753206  5.709077\n",
       "3 -0.575372 -0.929255\n",
       "4 -0.714855 -0.951365"
      ],
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.736426</td>\n",
       "      <td>5.697541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.936343</td>\n",
       "      <td>-0.931081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.753206</td>\n",
       "      <td>5.709077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.575372</td>\n",
       "      <td>-0.929255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.714855</td>\n",
       "      <td>-0.951365</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 198
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "source": [
    "mean_squared_error(Y_val_and_test, predicted)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "0.31270435"
      ]
     },
     "metadata": {},
     "execution_count": 199
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "This slighlty different (more hidden layers, more epochs) NN produced a MSE of 0.313 -- which is marginally better than the Linear Regression model above (0.33), and much better than the first Sequential NN Model which only had 1 layer (0.44). "
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "source": [
    "plt.plot(actual_v_pred_linear['actual']) # blue\n",
    "plt.plot(actual_v_pred_linear['pred']) # orange"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fad258c8ca0>]"
      ]
     },
     "metadata": {},
     "execution_count": 202
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAD5CAYAAAA6JL6mAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAACGMklEQVR4nO29abgs2VUduE7METnf6Q01qFCpNBmQgJIszKTJQhgxmjZDMxnc+sC0DbQnwP4au9u0sQ3Y3TINLiaDocEGg8G0EYNGaARSFUggUaWxSjW99+67Q94cYo44/eOcExmZGWNm3qlerO+rr967LzMybuSJFfvsvdfahFKKBg0aNGhweSGd9wk0aNCgQYP10BB5gwYNGlxyNETeoEGDBpccDZE3aNCgwSVHQ+QNGjRocMnREHmDBg0aXHIo6x6AEPIiAP8p9aPnA/hfKaX/Nu89Ozs79L777lv3oxs0aNDgjsIjjzxyQCndXfz52kROKf0wgJcDACFEBvAMgF8res99992Hhx9+eN2PbtCgQYM7CoSQT2b9fNOpldcB+DilNPPDGjRo0KDB5rFpIv8aAL+Y9Q+EkDcTQh4mhDx8+/btDX9sgwYNGty52BiRE0I0AF8K4Jez/p1S+hCl9EFK6YO7u0spngYNGjRosCI2GZF/EYA/oZTe2uAxGzRo0KBBCTZJ5F+LnLRKgwYNGjQ4PWyEyAkhFoC/CuBXN3G8Bg0aNGhQHWu3HwIApdQGsL2JYzVo0KBBg3polJ0NNo4PPnOCP3ny+LxPo0GDOwYNkTfYOP7lWx/DP/2ND533aTRocMdgI6mVBg3SGNoBho5/3qfRoMEdgyYib7BxnDgBRk543qfRoMEdg4bIG2wcIzfAyA0Qx8082AYNzgINkTfYKCilGDkBKAUmfhOVN2hwFmiIvMFGMfUjiED8xA7O92QaNLhD0BB5g41i5MzI+8RpiLxBg7NAQ+QNNoqRG2T+uUGDBqeHhsgbbBTpbpVRE5E3aHAmaIi8wUbRpFYaNDh7NETeYKOYS600veQNGpwJGiJvsFE0EXmDBmePhsgbbBQjl0XhHV1pip0NVsbQ9vHgP/+9xnytIhoib7BRjJwAliZjq601EXmDlfHs0MXBxMNjN8bnfSqXAg2RN9goRm6ArqGia6gNkTdYGU4QAQDGza6uEjY1IahPCPkVQshjhJBHCSGfvYnjNrh8GDkhuqaCnqk27YcNVobjCyJvCuZVsCkb2/8TwFsppV9FCNEAWBs6boNLBhGR90wVN06c8z6dBpcUIiJv6izVsDaRE0K6AD4fwDcDAKXUB9CYUd+hGLkB9joGuqaCk6b9sMGKmKVWmjVUBZtIrTwfwG0AP0MI+VNCyE8SQlqLLyKEvJkQ8jAh5OHbt29v4GMbXESMnBA9U0XXVJtoqsHKcHlqpUnPVcMmiFwB8JkAfoxS+hkApgC+Z/FFlNKHKKUPUkof3N3d3cDHNriIYKkVliP3wxguj6waNKgDm1sgNxF5NWyCyJ8G8DSl9I/5338FjNgbPAfw4Zvjyna0wou8a7KuFeBiRlTve+Ko6ai54HCCGECTI6+KtYmcUnoTwFOEkBfxH70OwF+se9wG54+njmx8yVv+AD/2ro9Xev3ECxFTJMVO4OKpO/0wxtf9xB/h//njJ8/7VBoUoMmR18Omulb+DoBf4B0rnwDwNzd03AbniH/zux+BH8W4NXIrvV6oOrumgu4FJXLHjxBEtIn0LjjcIMJduI2Re/W8T+VSYCN95JTS9/P896dTSr+cUnoqutrf+dBNvOVtH631nsdujvAlb/mDjQkLPnxzjB+vGKFeZjx6Y4Rfe/8zAICjabUmJJFGefkTP40Xf/CH2c8uGGHaAXvYiD7lBhcT8uQG3q1/F14RPNzMfq2AS6XsHL33F9D//e+v9Z5HH38Kn33z5/HU4XQj5/Dr738GP/hbjz3ni3g/9NsfxhU9wH/v/ABedvTWSu8RRH5t/90YPP12ABczIgfwnP/+LjtU5wAyobgLt5vZrxVwqYj8PvdRfCl9Z6337D77dnyf+ougBx/eyDmI3N2x/dxtlX/fE0d422O38HN7v4iXBh/C/c6fV3qfSK1owQkUbwjg4s3ttDmROw2RX2wETEzWw3QjefKnj238rZ99H/7s6eHax7qIuFRETlUTBjxENbZasWcDAHxnspFzEBHd8fRiEdSmQCnFD/7WY/iW1h/ihfssEjfCk0rvFRG56p+AuEP2swtWrBKReJNaudigAdtB98h0I51PH3jqBL/36D6+6sfeg19673Ov0H2piBxqCzoJ4Xhe5bdQny2IYFNE7vkw4T5nI/Lfe3Qfwyc/iO/DTwP3fR5udj4N7XhSKRXB8uEUkjsEiTxsadGFS62cZUS+P3Lx3sePTv1znpPwWQDWx2QjEfnUY8d48bUOvudX/xz/6Ff+jK1pSoGPvx14+uG1P+M8camInGgmAMCdVre2JAFbEJG3mRz5K45+A+/WvxvH0+eej0gUU/zbt/4ZHjJ/FLJuAV/5EwjNbfTJBMMKKZKRE8KEBxKz195tuBeuj1wQ+FnkyH/yDx7Ht/7s+079czaNOKb4L488jTCKz+0cSMhTK2S6kUaFKc+z/8w3vwL/82tegP/08JP4F//XW+D9+GuB//gVwG/83bU/4zxxyYicKf+9OtE1z7VtisgH7jPYJSeYnjz3Iq1f+9Nn8DeO/j3uj58A+fIfA7rXQMw+emRaqXNl5Aa4rs0ecNd078JF5M4ZRuT33XgrfjD6IVB6ubou/uTJY/zTX34P/ugTm1njH3zmBF/9799T6+EppYh8E51PIiLv6Ar+/v1P4/13/Wv8s/H34/DWk5j0XwLYB/UOeOtDwFu/j0X0FwCXishlvQ0A8OzqEbkUiojc3sg5SBFbYP7J/kaOd1HgBhEeeevP4ZuU3wV91XcAL/xCAIBkbaGHaaVU0sgJcF2f9ZxfVZ2LR+RnmCO/Z/x+/FXpEXjh+UW2qyC6/WG8X/+fQG9WK3KX4b2PH+GPHz/CjZNqegQAkDmRdzdU7Jz6Eb5A+RC0n/si4Oe/Ev3gAAev/pd4c+8h/NzBA4imh/VI+dHfBP7oRwFeCzpvXDIiZ+64vls9uhZPdpErXxcKP144qfkEv+D49Xf+Eb7HfwvGW58G8vp/mvxc7WyjQxwcj8uv38gNcE2b3ay7in3hip120n54+uQqhQ40EsFxqhMYohD4+DtO76QqQBo+BZlQkJOnNnK8IX+Y2zXaCGUeMPU3VOzsHf85flb5AeDkaeCLfxj4u3+CnVd/G/7zd3wBrN4OZBoCdTjC4bsVfzMB4rrYlLLzTCDrLLVSp3ApJ0S+mQuuxOymjKeHGznemeLx3wciHzB6gN4FjC6gdzEOCV70/303NAkwv/7nAEVL3mJ0tgEA9skBgOcVHn7khHiBOkut7CgORqMLFpFzMjmL1IrCycixxxh0lwxBs/Gx3wN+8auBb38PcOWlp3h2+Yj8zdaVRjZrEKizC1IjByCbaz+0JrxT5ev/C7D3ktnPNQW9rT1m9eccA3zXXwqHax6DhshrQzXYRQ7d6qkVcTNt6oKrnMiJfcmI/PZHgJ99U+Y/tSDj5Yjw5Kv/He7dev7cv5ndHQCAMyr/fUdugF15dp23pMkdnVoRUWWtmo6I9MbPnh+RcwLfFJFfv/0uPKL/M/zp5A8BbJW+nlLKAiYZsIiHqb2Be1cEctoyUUsWO6d4egipf0+149n8e2qIvD5UUxB59Y6RhHiDzXSZaLEHEED2Ltl0b/HgecMPADsPAO4I8E7wgY89iT/40OO4fv+n4yu+4BuW3ia32CIPxtWIfLs9u/n7mGLihQijGIp8MbJ4thfgV7R/ioeiN4HSLwQh5NQ+S6y9Wq2vghjs8yumx5z04g3tYnuTT8AiHuLxLQAvKH19EFEYdNZiHNnr32uE96VDW94ZKW2+6xwdoH1XxQM2qZXVofGIPPar3xjiZhJFz3WhUxcggOZfMiIXBHH3g8C9rwIAvOfjh/iGP/9jfPb9X4yf+cZXZL/P7AMAwmk5sYycEP3OFJBUQO+gA/Y9jd0Qg5ZW8u6zQexN8aD0Ebwn/gS8MIahyqf2WWLt1RKjCWKYnmMNRuSKN1RXEveKb1e7Do4fwSQzIqfOBu41sf5Vc+mfdJ4+nJ4coGJi5cJF5BcjTKoI3eJEXmPLpyVEvn5E7ocxDLAFZgbDtY93plhYyI8fTPFtP/8I7ttp4d993WfmR8zmAABAnWIij2OKsRugiyl7j7WFdsxu3ItknBV57JxacE+9l1znay+sUZyfReTnR+SU717phnaxus+UwaFXkciDCBZSor8NdIbIoYMYBFCMpX8ze2zQjTeqMbnsguXILxmRdwDU2/LplN1MIl+5Dhx/tsDa0cm5CiZqQ1wztYUTO8C3/of3QSLAT3/TKxLv8ExwIpfcYpn+1Gde5B06Zu8x+rAiVsu4SHlyyttQTXinXvDUeXogcOtE5Jz0z7EGI0R0ZEMkZYVDAEBU8YHmBBEM+IgJSxiUrb0qkEMHvmQAGam09oAReZX0IQAgjgBxThcktXKpiNzkEXmdLZ/BiVyJarSA5cAJIhh8y7dNxheKoErBb8pANvDtv/AInjq28e+/4UHcu20Vv0/vIQZJTLDyINoMW9GYpWPMAYxwBOCCETlfOxbxTr3gqfOHflwxEgUwi/DOMbUidq+b2MVSStGJ2TqoupN2/AgW8eCZVwAAqj9c+zyUyEYgLadVAKDfaWNKdUTTiikcZwiA95w/lyJyQsgThJA/J4S8nxByaqYFsmYipiRRa5aBUgodTMgi8pXrwPbDJCIfYIzjC+bsVwi+4P6P33sSf/jxQ/yLr/x0vPJTyjsIIEnwlA708KRQoShcDo2IR+RmH2rAopaRc4F6yXkEZZ1yRO6HMUy+VmqJ0fzzL3ZKfK3IGyDyiReiD7YzoxVrW07ArB58ixG5FozWPg81dhHKOURuaRiiPStglp5g6nXPJSLneA2l9OWU0gc3eMx5EAKH6JW3fEEYwwIj8I0QuRfC5A+GLTLG8DIZZ3GC+PlHbuNvv/p+fNVn3V39rWoPHTpJxDRZEHlwPTjhRD6A4jEiv0gRubjxzFPOkTtemBB5XKdoeAFy5CINuYl05NAOsEUEkVeNyNlDMGoxIreiCYI10ph+GMOgLiIlm8i7hoIT2oZUNRefLr42qZXV4EGv3IHiODZkwqJIfQNE7rk2JEIRSDq6xK6kdrwoGE9GCKiM1/2lu/H33/Ci8jekEOp99DEp9FtJW9gKIifeCQjiC0XkYu2w1Mrp1Thsd5qsvVpitAuQWpGjzaUjT6YOemD3SdUAzAlYagVGD77S5sZZq+/qbD9kOXclO41ICMFE6kCp2lKc2i15deofp4hNETkF8DuEkEcIIW/OegEh5M2EkIcJIQ/fvl2jOrwAlxiQwmoLzE21fWm0uvVtHoSww7WuAwCc4eWR6duTERzo+Nq/fC8kqV7vNDX66JFJod/KyA2hIIQcTACjDxh9EFBsyc6F6lqRgrMpdnrpVrs6RC5e6xyzoto5QOWR+CZ2sdOTA0j8gVaHyE14kDQLgdpFj0zWkulPvBAW8UBziBwAHKULPayYwkmlVsIN2WOvi00R+edQSj8TwBcB+A5CyOcvvoBS+hCf6/ng7u7uyh/kE2Om1iyBMNcKIcGEu7YLnegHjrssLeGOLo9xFvWncKDB0ur3TRNrC30UOyCOnABd8BuVR+QAcLdxsRwQRbqgBffMiLxW94cQroDOb+HPEMKGQnR8rQN3OAvaqhZPHZ+lpWS9hUjvrS3Tt3m3GdXyidxTezArEzn7XmyqJ3YG541NDV9+lv9/H8CvAXjlJo6bhUAyKufufIfdFGPShQF/bRc60UZG+vcCuFzGWbHvwKY6zBUEMEprCz1S7IA4cgP0CSevFJFf1y6WJ7kIAkziwT3FrhU/teUmdcRovg2A75jOKb2iJURebxpXFvzxjMjlitfBdV1oJIKit/hucD1P8okXwoQLUkDkod5HOx5Vc0C0jxBDwj7t19K0nCbWJnJCSIsQ0hF/BvAGAB9c97h5CGSj8pbPd1hEPlX6sODC9tbrnhDCDsL9GOLzVN/VBAmmcGCgpdcX8+qdbfQwxdEkPz01csKZ8yHvWgGAK5p7YSLyMIoTkjr1rhVndoPXauMLbKDLUnfn1Uuu8UjcJF4tx8IshJMZkVfdSYe8y0cxWyDmAD2s50k+5akVKUOeLxAbfSiIgCqdNc4RRqQNG8bGXFXXxSYi8isA/oAQ8gEA7wXw/1JKq41dXwGhbECLKy4ITryuNoBMKGxnvYse8ShLHrCInJxji1htBPbKqRWtsw2JUNij/N935Aa4porUSj+JyPcU+8JE5KyIxkkKfuKEeBpIi4BqtfEFNtDjxk3n1LkifE5MeJh66z3sKH8Y2VI7yb2XQdy3it6GZPWZle0aqZWpx3Puej6RE5O34lZJZ9lHOKFt2NArt0KfNtb2WqGUfgLAyzZwLpUQyhbUuFrhMuTbHl/fAqYib1mhdzoHoh9Y3WJELruXh8ilgKVWViFyiZOyV6B8GzkBXqC5QAhG4txlbke2MbIvRh85K6Kx9JBEaBL5nQbSzoF12viob+NZ7OAu4Nwicp0yYzgLHp5d82En8WDnRLsC1at2HYSNAtEsqC022GSdYGDqBrDgYWrkO6lILea34o4OYPDUaR6oc4zDuAWb6hvzo1kXl679MFbMRK1ZhpBHRdRkX5JX0bQnD2Ibpbb6cKTWpTLOkkIHDnRY2grPbk7kRRLmOQvbVGplIE0vTGrF8aOktxuo7v2xCkRU6UGr3sYXxyChg//6OL8tz8HznlIKEyK14mPqrqeVkN1jONDhqAMYFXfSSQFRtaC2t6GTAPZ09e/K8RwoJIZSQORamwV4k2F5R100PcQRbcOFVq/+cYq4lESuo1pEnhQiWtwUZ83USpxaYLbSgxGs7wFxVlAiBx7RIddsPQSQEHlcoHwbOSG2ZX59jR6g6IDaQp8wIr8IcyuT/mQOWsfMqibEWhlJvcqpQNFDfkJbCNX2uaRWPM9jU40IE8840/Wuke4fYyL1ECsWdOpVWgfCDweqBcnqAwDCqvL5DAjXRdXIT61oXcYR9rC8E43aRxiiAxt65Vbo08alI3KqWDAr9oSLm0npsC8pXLd5X7SRaS146gDtaHghCKoK5MjJ9ZooBSdyUpA/ZF0rNqD3AImnb8w+OnSCKKaFqtCzgu3Pu+rVsUOuC8qDCFvpJwXWUvD1ZcOApw7OJbUitBe20md/X3MXawRD2EoPVLVgwq3UOUaT+8ximgQA8Rr1KOEHrxZE5K0eG6Dij8s/R3KPMaQstVK1E+e0cemIHKoJnQQIgvLtukiFGP09AECwZgSWDKdQTYTGAH2MMVmzE+asUOQ1UQpO5HKBcdbICdAjkySlIt7XphfHAZGlVmakuqnxf1kQZGSrg6QLpBR8vTpUx1Ttn0v7oUg/uhqvizjVp3FlwYpO4Kp9QLXYtJ8q94vIO6tmsp7W8SQPPPY7kIKulQ53QAzL0lmBCzl0cEw7cKFvRP26CVw+IudfhmNXWGD8RjW6zLNh3Yg8EXaoFmJzi/utnD9BlYJSaLGDSF72Yq4EfjNp/ghxRl9xHFOMvRCdeJKQPoA5K9uLoO4UrnqxxIdcnKaYgz/0XW2Q2NlWfY8NHROpez6pFSGiM1jO2F8zHdmNR/D1AajWggmv2s5MdPmordlucA0r28Q+t6CPvN9pY0KN8lm8/IEyBOtaUaqmzU4Zl47IRVN/pcJlYCOgMtR2HwAQrVlhliIHARRAVkGsbWxhXCiSuTCIfMiIc02DSiGr8OUWuphkKuwmfghKgVY8WojI+4ns+eQCPPBsLv2OePE7Gf91CiCBjRASArVbuTgvUisOdAxJ71wcEH1OerHJUg3rFIT9MEYfY0TGFojWQquilkNK7XxFaqVoN1iGpFamFhC5pXIHxJLP4XWiIW0jlk3INALC8+eAS0fkUg0iJ6EDl+gwzPqThbIghw58wqJapbMDi3gYnlyCgicniDzToEqH0HrokwmOMh5cojXMjBYicnMAzb84DoiukGq3GEmdZscBCR140EFVCwaqFfnSqZVD2mGplTOuwYhdq9zmRL5GOvJkYqNLbFBrC5LegkwonAoRfiKg0lpJYKCuYWWbuE8WpFZUWcIY7fIHBo/Ifa0PKh4MF8DK9tIRucyb+qvk7qTQgQsDxgqThTI/O3LgSzoAQOUFVOdkdQOwMwP/vWlBRFKG2Oijl+OAKPzG9fBkgcj7kLmV7TqCjk3B9pnCT3QxSaco5pBCGx4xQFULGongeRXSK0mxU8dB3AYi78z7lJM6En/YRWsEPxPeASK1dlL3bXkAJkepsYR8sIkICFYBTXWbFWEqd8uHWIhdkjWY7XAbIq8PmbcQVZlMLoc2PMmYKbrWvCnUyE06P8weK6C6l4HI010Aq8IcoE+mOM4icjcAQNnNthCRS5ELHf6FiMidgPlcS2323Z1mx4EcufAlPbWDrFLT4Z0u0HEj4B0WZ5wnF/lkqc0edusQuT28BQBQ2jtQTBZMVRnALIfubL6mJMGTOzCi8eodYn619e8q3fKWYp5akVvboILIL4Bx1qUjckVnCzyssMDkyIVP9NmTeM0ITIldhLxgaA1YATW4DMZZQbWIpAiytcU8yXNSK204IDRKcpoAElLv4WKIghyPKfxkTlJqdHo3oBK58CUzqek4FWs6AEutPOPz4OOMRUEhf5hIPCJfx0tEBDl6dxdKEoCVp0jU2EWYmq/pq110MYEbrGZ6l8wvUPNTKwDga72kOJ8LHpFr7W3ETWpldWg8312llVAVvdOKhhDy2sNk9VQLnywW+mUgch4xFHlNlEHtbDMHxIyI/MQJkuEBi10rAHDduBgOiIHPBoMQa4AIcmUTp1WgRg7C1G7Qr9LGl+paecrl0d4Z95ILkZTCBTJ0jeAn4M6HZm8POo/Iy+5byjusglSrbKh11zLOSoi8JCIPdd4uWxT5O8fwoKLd6aYCxIbIa0M12Y1RZcunxi4CHkF70EHWmEFIKYVGXcRiO2WxzofKc/7OESKqktdIrSjWgOfIl3O9IzdEn2QQ+QWzsk0Kd2qrlovmKlD4Q1847vlVBhCIYid03IzOJ7USc1LSOiz9RNaIyKMJewi1BntQeQBW1gLsRzF0eHOtsrHeW8vKNjEtK9mRUuGA6OU/dGP7CMe0jZ32LG3WEPkKEBE5rUjkIoL2JGOtYbJ+xPKrCZGbfcQgkC8BkYsoSC5QtpWBWANoJMJ0vJxDTMRAwFKxEwCuqhdjSlDStaRZCCUTWuxm9sVvAnrsIpINKDwiryRGC2zEIPCJimPKItizFgWJwqDOa0BkHQk63010tq7AaHUBlHeOuX4MCx6iVEROuZXtyQpDvKOY8lSNPlMc54Dw4CyY5O+CgskBjmkb2219JjBqcuT1ITpQqkTkOnWTBeETfa1hsq4fszYyQeSSjKnUuRTGWaIwrKyRWhEE7Wcs8pEb4KoqvMj7S+/ZVZ0LkSNP2tBUC5FiwiLe2sNG8qBR9tCvGokCAHwbDnRc6ZgYw0QsqWcv0/eniCmBbnYQELXyfNwsyM4RJjAhq3pS24pL+tKFQ2Wc0jwQc/XhErYfwoJXSdWsVDDOiiZHOEEbWy0Nst5E5CtDEHmV3J1OZxF0IJlryWntgM/9SxVMbKUPMxiufMyzgu+yraJIS60ETsrhZHkHMnJCXFGdudel/7wj2xeCyJM2NK2FSLFObbgEpRQGXMSKlRT5KvVjB1PYVMddAxMAQaBvnXlqhYQOHGiQZIlN41pjF6t4RxgRFomL/HRZ8dT2Q5jEmxUSAcgWi8jHK6yhqcfUvFEFItfaLCK3ixwQnWMekWuQhAlXQ+T1ofOLVyV3Z6Ry2uvmRG2fRQpEmy0IT+ujFV18QVDoch91o7P6QQRBZyjfRm6AXSUjR651ACJhIE2TXvPzRLqfOFZMWKc0t9MLY5jwQVUzSQVW2UHGvs2IvM/WmKv2z1zdSQIbHmFaiUAy15Kg6/4QE6nH/qJWS0OIwcvpQclqewsqiWBP64uCxJi3WC0ncoMbZzmj/Ien5B4nOXJxP12EcW8bI3JCiEwI+VNCyG9u6phZkGSJGbqXPAVZVOQnBY5INqvbiWZAeFmnjXdCfYAeHcELz9/ZrwhiW69bq+fIZ54Xy6mkkRNgW7ZZ32/6hpEkwOijTyYXIiKfdS+0ECvMxMk5BVdGl5MRVAuaWf1mj7wpbIiInI0oPOscuRS6cMEKjZFsQI0qqlIzYIZDOCon8oqFQTdg6tt0wKR1WKTsZ+wGyyBSK7SCqrnVY506/jjnmlMKzR8mqRVNtFSu66q6AWwyIv9OAI9u8Hi5cEl5B4rnuVBJlCygSDagVTUvyvpMfjwp1fkRm1sYXALjrMibIqAyTGNFrxUgIXItOEEYzeeVR26IAbHne8iT9/XRpRM4QQT/lPLRVZE2PRMmTu4pROS268MgAYhmJfYQVfqxY28KBzqu9wwQwrzMzzq1Ikd2ol6OFBMGXPjRat9bOxqxXQUAKAYiSKU5d8ePYRB/Tk6vtdjai1boqZ/weZ1VNBTdLS6CmuY8MPwJJBpiiDYGlgbdMBBQeS0bg01hI0ROCLkbwBcD+MlNHK8MHvTS3J075S1E/AuMVQt6VfOirONxQYecKhgSawcDjHGc0ZJ3kUB5EW2VMW8JOJH3McVwIboeOQH6ZMFnJfW+Vsyu3Xl3rsiirqJZzFb1lFIrwk+EaBaMVvWaDvWncKiOjqGiZ6oYonPmxU4lcuFxP6FYMWHCX3luZ4eOWJ4fAAiBR4xSNS2LoN25gImYgsiHtc9BzOsssrAVGHRaGFMTNC+dxX/uq33IEoGlyXCgVRInnjY2FZH/WwD/EEDuo5sQ8mZCyMOEkIdv315P1u5JBqSSDhRhkC8WBFVMGPCXosmqEJ0fcmrKiNzehk5CnAwvdgti7LMte0tfg8hVE5Gko0cmS6KgkRugQ/OJXKjlzju9MvPwaIFoLZinlFrxuRxf1izouoGQSpVqOtS3YUOHqcnYsjQc0A7gngDR2V03hQuZAObNU9lDfAE0cNCCi1gMNQbgS+VELrpW5ibei04od1j7PERqhejlEbmpyjhBCyTvc3ircWQM+OsVONDXsjHYFNYmckLImwDsU0ofKXodpfQhSumDlNIHd3d31/pMn5hQSiJyoaQTyjrKIzB7xQhM5MHSc/9m46EuuN9KYMOhGsxV5nWmEOp9JtNPEXkcU0y8EO14nE3kRp+ZaQHnLgpKlJyaBUlvnVrXihgpKOstEEmCi/KaDsBsdR3oaGkK+paK24ko6Oyiciai4yk47tw4XWEA81TcE9aMyAPJKFXTun4Ak/hQ0sSb1GeGtc9j4rEuGFkvrw8RQjCROvkOiDwil/jvZGkybKqf6oCSqthERP45AL6UEPIEgF8C8FpCyM9v4Li5CCQdaknh0k/dTAD4ovRXjsBEwTA9LsrsMb8Vb1Q+5+9c4TtwYMBS14jIwZRvfTKd82Afe8yL3IzG8z3kAuYAGrcgPc+IPObCEACAYkLSWzDhw/U3f04BDyKEAMsl+syatQAkcGBTHpG3tJRx1tkRuZYS0RGNtWiuklqZHjPDLGFlAQCBbJV2jgW8w2pOvMZrL0qZM2EGbJ5akQvmdc69Xu5By/scbmErt1jx1dJkuNDX8qPZFNYmckrp91JK76aU3gfgawC8nVL69WufWQEC2SxfEAupEEmzoJMQtrtaPlsUNLTUgrAGTP0W5lW5LwikkG3ZrXVSK2CRSJ9McDSdkZ+IspcsbAXMAWTvBATxuRI5awn0EEjMUU/RW5AIXXsCThaEilMM+/UqErkUOiwi12UMLA1P+zwqPcPOFY3O5PGSZsEk/kqpFTHEWO3Odt+hbEIvCcCCJGBKEa/eQQQJqr9a+6EFD2qFiBwAPKULM8z5HE7kWpc9nExNhg0d8M9/StCl6yMHeAdKCZEvRtAS36q5VexEMyDax9ItfDpfpNEFN86SQxsuNGjyel+33GJ+K+mIfOQG0BBAjZyciLwPAooOnHP1JBdFNCEMUXgP8GkUqsQxNb72PFKeUgDY92RDh6UqGLQ0POmcvXGWQV1EvFVPSsaz1f/e/NHM+VAgUkzoJfetsNFV0kROCBy5A32F4RK267EOooqqZl/rs0lXWefGu2ZMTuSWpsCmOkj4HIjI06CUvpNS+qZNHjMLkWKVthKGnhDB8IicP5Gr+CFnfibPg83J3DdgnPX2x27h/37nx1b3Wq4AKXLhSwYItwVdFbK1hT6x53LkIyfMdj4UEFa2ZHKuOXLbj2ASLxkGUEs6XxOCjDSupPUls9weIgog0RAOZTungaXhZnj2qRU9ZUOhGIzIJys8gEMe3IhdK8AmVBlwC9tQxThGstAu6CldGGUWsxnw3XoWzrHRZ4X7ePkcvdFtjKmJQZd9ryK1Qk5xQElVXMqIPFbM0jmIEb9BhbJOFE/8VW/clLw7gd5FCBmKuzqR/9J7n8K/euuH8R//6JMrH6MMSuTAl9boIRcwmbgn3bUycgN0hfNhVh85/9kV5Xxl+kJoEicktZnxf1kQD32xewskA2qZPUTK+bClKRhYKo7BifysUiuxUKSye0Ux2lBIDNerT1TC+bDTnxE5VXmBuaBOFWfdZ2Ce5G06rm1yFnJ7ispDVcwBZMSg3nJU7o+PMKRt7LTY8G6RWlnHxmBTeM4SuTBI0i3m9SBu3Ep2ohmYybtThEgIJlIPqre6cZYgt//tv/0F/vgTpxN5MW/sTRD5ACY8jKezazhyAvSR4XyYeg8AXNW8c4/ILcw8PERf8WkUqsQxhQd3KBulxfn04GVDlTBoaYggI9R7ZxaRJ2kmvsbFjqLSoPNFOIc4oRZ6rdS6Uy1YxIUd5Ef48cI5JOem9dDFFOOa+frZ8aqlVqQWN87KmPwV24c45qpOALBUGQ7V1jLj2xQuJZFDtWCQADQuf7LrPCJXE/OiFVuFwuwtmq30YKxhnHXiBPjLn7KFe7csfMf/8yd4drj5RaGmJhuthcQBcbYDYV7kVYj8fB0QnYClVpLvj0dosbf51jGx9hLDLNksVxXz94SyCUJIQha+dnbGWck8TX5thGNhlbGKi1DcIxyjC0NNUQzvgrELIvKZ+naeeNnM2PoOiJE3/zuVnjfvSJkcLRM5cVhEvt1myldFlnj94/R87avichK5VmGQK78xjBZbjMLzYtWcKMkZl+auaZw1cgLcPbDw0Dd+Ftwgxrf//COblY1TCp0yJ761wUk5TkmYT5wge6hE8p4+AGDvnFMrDo/Iky17Mt1l8xE5SdnlAsznp1RVzNeXKDQOLBUA4Kj9M0utCPWyENGJMXXBCukn1TvGWOrO1WUkvQ0TPuyCnHvmzhfgab365mvJg7piRK5zXxf7ZPmaS+4QQ7Sx09aSn4XyesZim8KlJHLCv+TCLV9gI6QSNJU9PYVr4qoqLDl0EEECFH3u56G+hW48QrTigIITJ0DPVPGCvQ5+5G+8DB94+gT/5L9+cHPFz8iHjHg28XsdJA6Is1TSyAmwl1jY9pffw3Pk2/L5DpcQrnqCnBJCP40eYFH84us0Vq3SVODigOyBxchiKvfPzAFRpB3JwsMuWiH40f0hJnJv7mcyb/l0CwKwvLFsktlnqRWnZvuweFBXjMgNbpzljpYjcs0/wRAddA01+VkkG1BoeKbq2yxcSiIXas2iVkISMH8RIrFfUWutZzkphQ4b5LzQ+SGMs1bJ/wZRjKkfoWeyhfGGv3QVf/d1D+BXHnl6c8XPICfCWQWcyI1wlOwaRm6APdkGQAC9t/we1QAUE1vS+Q5gFl0rksYLiJyspFPwkiahAw9aMpGGKiaM0tTKfBTfM1UQApxI3TNLrYhGgKQzi6+ZaAXlYisczgyzOETnmFd432aPZVOsLciEws6YUFWEvJ10Htp9RuRLU4LiCHo0hqf2IEkzDghFgHTOnuSXlMh5B0rBk52EDlwyywsLIU+84gVnnR/LeWZibXPjrPp5sjHfYnbNmXT+u173AF7/kr3NFT/5TUgrbi0LIYyzyCRxfBw5IbYVm0XjUs5y4qO6ztOT3OFdK2LtiBt7nQk4eZBCJ/H0Bpg9hEoi0HB5cHWCBcJRZAldQ8WxMM46xfbU5BQW/YT4w6528EMp2vEInjafahMtn0X3bVI4XCBelU/v8QrGsGVBynkw5KGfOCAuNDA4Q0igCPX+3I+TlOU5y/QvKZGLBZG/wOSFm6mqsX0elMhlqsDFc2lvQyYUo2H9qElEqCIiBwBJIviRr355Uvy8cbJm/o0TBFlj8HIC0ROOadJLPnIDbOVZ2Kbe18UYIzc4tRmZZXD8ECa8mVeOiMhPoXVMFrs3AbVC66svvqfZA3erpeEg7gJxyMyzThlhTkReW7kY2NDhI9TniXzWOZYfkUuicLiwg9Q6jMiDGp7klNLcVE0eui0LY2qCLGpDeDqRGvO/E20i8tWhVjB0ZzdTinjFwlixuJUe5JyG1uHGWcf1/VayiBwAuoaKh77xs3Aw8fFrf/rMCmebghBYbILI9Q4okVkvOVd3JoOXswqdAmYfrXgMSlG7fWxTcD0fOglnRC5riCCVuvGtAiVe2L2J4vy0qKbDvicpZRY1sFTcOkNRUCKi45GzCH5oXZLi50pTzofArK++KOeuhg4CMktLCQg1ZdaowTyIugg7cLX1L0kEI9KGtGjQxYldDGgWoBUHZpw2LieR6+VzEOXFm4kT+aoqLDaJezkiN/vMOMs9uVX7mHlEDgAv2OvAUCWcrDm0QvQGyxX8mEtBCKIFB8SxG6KbZ2ErYA6YqRbOzwFRRJuJPSoh8CWTWQtsGEo0v1bEkN7CLiu+LpWUJ8jA0nAj4ERxBkQeJyK6+Yic1CSpYMzOlbTmSU/jmo6wYACzknZfTEHlwyVoDQdE4UXOPrz6+p9IXSiLQ9V5wVlpz/9OiQK1Sa3Uh4iqihaEupgKIQQuqpkXLYLyFr6szo91jLOKiBwAOoa6dqeHl+GjvhbMeQfEkROgTXMsbAWMPozwfB0Q44x+4kAyoKwxxzUPauzM7d7Ew8OzC7xCRO95akD2oKXhae/sjLNE/7to1RXXqmwa1yKmQ+582N6Z+7nwnqE5OXdKKbTYyQyYsjqmSs+DW9jGRAZkrfwNHI7cXfJ1CbjlgNFdIHLxgDiFNtY6uJRELrZocUFEri3cTAAzL1pFheVHMYyUB0UaLS5BXmUMVRmRdw1l7QKhKGApGyJy2dpCj0fkUUwx9kI2OCKr9VDA7CfOdecVkQv/k3Q/cSCb0OPNR1Ja7M2tPWGl7BeNBAum8KkMQ5+R2FZLwydd/vcz6FwRamgjSa3wwmtNIne5KlKkHQVISfGU3Wf+rBMkDb6+pBq1AuF8GMnmUrdZEXy1t+SAKAYyi/ZEgWSHd85+K5ebyAu2M1rsJU53AmxCSf0L7izIu9MgFo86Vtj6ClLr5hG5uX5ELjoE0j7q64BYA2zLUxxPfUzcEAQx9LAkIjcHkCMHOvxz6yWfeXjMvsNQtqDFmx/Tp9GZpwuQUngW5IZjf7o0jq9vqXj2DD3JKS9qJg6fsoYYMuTYrVWk9sesXmSkfFYAJNc+zxbB9ZnVcJyRWoFqIYAC2a9O5DYfmJ513xYh0PtoLzgguicHiChBtz8fkYu0WZNaWQHCw6LIJyMrFcImlNTfStt+xAbCZi0IzYILbSXjrJETQFMkGDkDH7qGurb1a+LvbG6GyGEOmAOiHbARb3AgIS4tdgJAF+fYS77Qpw3w4cLUXVnMlQVKKQy4C0QunBbz12vkTmHDmCPyLUuDA4Md6wxSKySYwqMqTJ2nIQhBKBtMVl9DbRyNDxFRgtZCGkLkqfNy7nbAOovirIicENhSOxlSUgVTPniZ1iRymuGAGEwOcIIWdjrzaR+RsiwKKs8Cl5LITS67Lxpoq2csCDaQYoWInPcg51W+VzXOEqrOPHRNFeM1iU8oWbUNEnkPzAHxRHSs8J8XvQcAegUS69/50E1800+/9/TaEzOEIZFiwiLeRi0R2AALHzTVPieufVTg6xJ5U9hUh5Uax9fn6s5QH5yNujNwYEOHmvKtjxSTeZLX6Dai9gGGaKPfWsh1q8VE7vhRIfE6Shd63tCHDEz5vM6qHSsJuPjITQ17pvYxjmkHW615Zbd4SBd10J0FNjGz0yCEvJcQ8gFCyIcIIf9sEydWBE1R4FANpOApaFB3aUFEsgF1ha204y/IuxewqnFWKZEbytqpCJEb1q3OWsdJYA7QolMcT5jkPvEiL+kjB5Cr7oxiin/xW4/hXR+5fXrticFyaoUqFswNz+0U2/k0eWiJPURBasVbTq0kxln64Exy5CS02XzR9HnJJkziYVLjeyH2EY5pB31rYW0rGkLIkHIGMThBBAM5O18wT3IrqheRm3DnevOrQIxyGx3NOtGIc4Qh2thuzxdNxU43cFcbWLMpbCIi9wC8llL6MgAvB/BGQsirNnDcXBDegYKcHuA48KGRaJnIFavcvCgDtuvDIEFiJrQIVx2sZJxVRuQdQ8XICdfyXYl9VkSzjA24HwIJKYfTY4ycsNgwS4CT/PUcB8Tf+dBNPH7AjnNaxdCkXzxV7BQDuVed45oFITxKD0YQ9hBFqUCWI9fmIvKtFlsbtjI4k9SKHLrzIjownxgTfqFj4dJxvGMco42Osby2PWLkdo4xz3g3104iULuw4uqR78RjEb5UcTqQgFCRTo5nfiuyN8QJOujo8wPMTd1ASKXVXVU3hE3M7KSUUnF1Vf7fqcv3igbaiinmi1NGYsWATr3axCgUeXLOggiMATrxqPZxy1MrCvwohlcwUaUMsc+irJamlL+4CjhhU+eo3It84T17qru0w6CU4sff9fHk76eVQ0+k2umHsdaCueHUiuM6UEg8N1rM4DWdwjyqb8OhbF6ngEitjOXumRQ75dCGRxbTISYsuLUicp07H8rScqdIIBm5XTDMD8fPjaAjvYcunRROGJo7nsceqnn3bR70DmtgcEezh6ceDOEo3aUpW5auwIG+krHYJrGRHDkhRCaEvB/APoDfpZT+ccZr3kwIeZgQ8vDt28vOYnXhESNXlefwft3FVAhVLH7j1iNG0fmRtyBiYwsDjGpFLUCV1Ar7t7WiVJ/NgTS19QYvJ+Ck3IrGuDly0asSkQsrW3XZyvY9nzjEB54+wZe+7DoAJB4um4YcZZgnqS20Npxa8RasYAHANExElBR3NgT2UrGzz9fGCTmb4RJylg2FasEgfq25nUY4hC33M//Nl63cOpVIYebtfGOjjx6p7kk+8UO0iDdrEayI1oC1GHopbYgZjuCry6ZwpsaIPD4NF80a2AiRU0ojSunLAdwN4JWEkE/NeM1DlNIHKaUP7u7uLh2jLnzJgJzTgSJEMEsLQjVXGibrJ73Y2QVD0tpGj9g4Htf7MqsUOwGs17kS2LAXIr21kJrB+cThNDVUop//Hr0HgGBbWk6t/Pi7PoGdtoZv/dxPAXB6EXky/Dh1UxO9BYt4cLzNfaaX8dA3NBk2jEKFpBTwwcupnZMiS+iZKo5pB/AnQHC6AwzU2F0aCSjxYRATr+LDjlK0wpMl50OBUDah5Wg5HF6clPI0D2YfPWJjZFerc9k8tVJ5zBvHzAGRF5hDDwZ1ERrLwYqlybCpPvNRPydsevjyEMA7Abxxk8fNgi/lTyYXIpil3JjWKp1QkgUhc1fN7AUmiiPjjKkieYhjiokXomvkpzzEv61T8JS4na+hbJbI+5jik4c29hSbRbkLPu3zJyEBZh9b0nRud/EXz47w7o/cxt/8nE/BXpe9f+gUOASuCEop1MhBRBRAThmU8fXhFQl1asLPUNKKmk6RQlKKHDZ4eWHnNLBU7Me8UH3KUTnzE5r/HiW9Va9rxRtDQYhAz96hRbIJLadO5XsOJEKh5ETQEl979qhaB8/UE10r9SLyHndApOJ6i46hBe8YADBVGQ60y0/khJBdQkif/9kE8HoAj6173DKEkpHrkyHc1eSFCJpoFkziw/Frjotyi1v41B4TPghpchWM3RCU5ouBgFREvkaUKixVpYx85UpIWdl+8nCKHdkuTqsIGH30yAQnqfbDf//uj6Olyfj6v/w89E2WDz6NiNyPYujwlzw8RNQcrjjHNQsBXyuqsTBJiuiFlrly6CxF5ACT6d8KORGdcudKlhpa1lv1ulY4+YXGMukBs979rDbTMhWyyudpuuNqD7SJG8CAV9uLX9d0jKgFOEP2A24LIFnZEbkD/Tmh7LwG4B2EkD8D8D6wHPlvbuC4hQhlE2qOWX/giJtpfkHMPC/qRWCi/1fLSa2YBVNF8lAmzwdSOfI1Uityjo/6yjBYnrBPJjiY+BhIFYncHKBDJ8nu4qkjG7/5ZzfwdX/5XvQsFYYqQZOlUyFyocxdVPqeRg+wUG8uKmlZTSfnZqcUSuTAgbaUAtuyNDzrn41xlkazrlGrXtcKj16plU3ksdrKbflMdr55RC48yUfVrkPgO5BAa6dWAGAsdSC7jMDFfa10dpZeZ2kKbKqDnIKLZh2s3cpAKf0zAJ+xgXOphVA2oecYHokFsZjTFnLaolFTWYiS42UvMIs7IIYbJ3KeWlmD3NhAjPVrEgkkGVTvosd7gftkWtxDLmD20R7fgh/GcIMIP/n7n4BEgG/huXFCCHqWurbbYxacIIJF3GQepkAVxWVd5D30fZJf00HogoDCocZSCqxvaXjqGWGcdbpEbmSooUWOfFqxrkTtAxAA0oLzYfLvqoUWcTH1Q7QWWvnEzjevFqULT/JptdRKYomwwlCVqdSF5g8BAJPjfRgAzM7yfWTyiFxqTLNWQ6zkD7RN1IwLxJtEYAXG9pkoGRfVGTAij2r0+lYi8qTYuQ6RZ9vvrgVzgAEvcvbopLjQmXqPsLJ9/GCK//TwU/iyl9+Fa70ZcfRM9VQi8sRzY4GkBNkWCXXqQqy9RQFWUFDTEd0soWwspcC2WiqecEVEfoqpFUqhw182hlNb0EkA261WYBQte2o7O3ggGhdhZUT44iGYJ7yzeiwijuxqKuosf52qcNWZitThg5jNXlZEzoh8FTO+TeISE7nF8l9Z/5YjSxcRdaELXQZohk/H3HG5XSepsfVNiHxR/ZaCrrB0w3iN1IqWMxBjHRBzgF2F3SSlFrYC5iCxsn3L2z8KN4jxbV/w/LmXnBaRi9TKokBM5WSbZ6u6CsTa0xfWXiDn13SEBWqWTXLf0nAzMECJdLqpldCDjHhZHs/zy6FTLXXgnTDDLK2bR+Ss4WCa0QUTl9xnFh8uEdvDSueSfK91JfoAArXHXD0BeNzCtru1t/Q6VZbgET3/IX1GuLRELloJ08Y2Aokd50JUpK6aEy0b4KpomMCCVMM4q0pETghB11TWSq1o1MkkiLVgDlhuHOAWttWKnWowAkGM//7nN/H6l1zBC/bmv5++qZ5KHzlLrWQQufDH3mAPsJimI+vznxXIJtQ8VTEvlMXKcgpgq6WBQkJsnK66MzdY0SqMqUvBHx8gpBKsbnaOnOht1nDgLQdhySSivK4VXmwkbrWIPOkkWWGoSqj30aEs8AjHh/Coiq1+P/O1/opmfJvEpSVycVPGGdVi8QXqrfmoSEw+iWrKaTNVgQsYS91axlkiXdLNkDGnsZYDIqXQaY6b3DowB+hjAh0+KzhXTK0QGqMNtuC//dXPX3rJaUbkzCtn/oYWf99kRJ6IfjJ8fvJqOsKZMc7orhhwdWdwyn4rHifqpbSGuM8qttfFkwMco5OoUhchOoVce/nBQESeOa/LRDXgQYPkVbPDSDxdVojIqTFAl04RRRGoc4RjtLHVzm6xDSUL6ikMKKmDS0vkglRdOyPf7duIKIG50AKmJy509W7cpP+3gBBtpV/LOOvECaBIZKlveBEdU109Io98tl1W6i/kQvAOlMQwq2JqBWBColfcN8BnPW85Yuuu87sWwOaplSWSOo15izlT2+Min59kx5dF5OxB72lbp+qA6GcoUtk5sb9XvmfsQxzRTm7KUNSpssbeCT/0IuKdkDaUCp7kXhjNyHWFiFyyBpAIxWh4CMk5xgjt3Hs1UgwoNACi85lHC1xiIhcLzst4siNw4ECHutABIIpbeRNKcj+L92JDyr9cntpHK6xunCVUnYveDYtYywGRR3p0E4OX0zAHaMVjDMg4+Xv5e/oAgFffo+AfvvHFmS/pWyrGXogwWt1bJgtuEMHMkmqrmydyEtiIIC0JpFhNJ0fsxKPdLCm5cECcKv1TTa14eSK6ZCZltXuGOEcsIjezI/JZp9ByACaVpTABTOVOJU/yqcetp0uOlwcxpm50dAuKP8REWvZZEUh2vOc4gPkSEznvCc/oQJFCGy5Z3gaJRVo3J6pEDvxFM6EFBDozzqqKMnm+QNdYI0rl0aG0wkIuhDmAhBh3kYPk71XeAwD//Avvxivuy86f9jZhSZABEZEveeXwNVQk1KkLKXLgQV8aLUZVEypCIMr4LkVKIYPIB5zIx9LpGmcFeX5CfJeQRMslUL1jHNP2soUthwimfGf5HpSi8hSmK3eWxrBlQQyVKDteHrQOa5+0h7dhBCdwM3xWBJIdb0Pk9ZHMQczYokmhDRcZxMsJrWggRRaUyFk2E1pAbG6hj1FlZ7aRExSqOgW6prIysYkHVl4718rgpHwf4UrWKn3k4jUFw3MFkW86T+74ASySQeSyhggS5A32AEthjgBLrD0/YwfpiwLpcv+0MM4aki7gHGUW9zcB0cm1/LDja6fiw073j3FCurlTr7RWFwAy3QITwVRB4OGr3aSNtQhTbifMjlc/tSJaDZ3RAazwBL7Wz31tUts4R+OsS0/kQUYroRS68DMi8lVzokrsIpBLerGtbbSIh+GoWnqlTkRe1e1tEWK7TDIIYi1wIn8euTn39yrvgTvMfYmI4ob2Zv1WxBpRFi0WCIFLzI32ALPdW8ba4+QUZBXa+XrM8s1WZAldQ8ER7QI0rjVFvg4EsS6Jcfh5Fxl+JYhjGOEIjpIfvSbTkjKIXIpchJj3w1lEoHXRouUdNHOplRUCmRZXa/vjA7TpBLHez39xkqI7vxbES0vk4qYMMiJyJXLgSRmFyTqLMgU9dpeky4uQeE5tfLRf6ZiVidxU4QYxvLC+1aooBCv66UTkf/WqPff34vf02f/PISIXgp8sG2KvSKizApQsK1jMdkVeVnGer8c8afpWS8N+dLpDmEMuxlkyhuP3jE7d8t2mdwIZETwtfz2I7yDL9lWtsPON9D46dFrq/T+XWlmha6vDrWyl0dMsJZZjOQCkdrxNaqU+xKIPMwqXTM2YERVJMnyouQMpskAphZrhQbEIIYCYDqsTedcsd0gQMv1VREG5ec91wUn5enwDIDKgVxgjp5qAYpQQ+ekYZwkJftbAglUHcuchT4Al6SI3nNGtUeIx0rc03Dxl46yZGjo7IjdQwZOcd9XkOR+y4+W3fKoVdr5U76NDHDhu8Xc29VKDnAuaFPLQ4qkVZfg4AEBuFRE5/26a1Ep9iC1aVgeKGjtLTncCPqknp2XDdN3SXmyDb8WEsq0IlFKMakTkwGp+KyK1opobmtcpICLw4ZOM1Es6b+beJxzlMnBaEXmRMjeQTGgrDOTOgxq7iDLISOatsFlEHnhTOFSDpWevh62WhmeEcdYpda6ICHnJ4ZNHm8yTvIzI2W4hynE+ZMfjRL5Ql6CUQquw8yU8iBgPi3cmU5+Njcsb5FwGIisYoYXW5AkAgNpelucLJCmxJrVSHxonp6z+1iJZui+Z+S50GWBiEh9xyYJoceMsf1x+o028EDEtVnUKdBJP8voReeLEl+OjvjJE4TIOq6VV0u+rklrZsLozmV6fEZGHsgltg2IOPcN4CkDisZ1F5KE7YRa2evYOrW+peDLxWzmd1Arl12hpSDf/XUxSwcefnxvJMcwCkDwYFgens4CpfOcr1J1OiQPi1AthEm+l1kOBidTBjv8MAMDI8FkRkLX8dNFZ4dISuWnxiDxDcabR7KgI4FvpGjeuE/CiSYmncWfrKgAgmpQTeRV5vsA6495ESkHfdESuGrMbpA6Rl0TkmiLB0mQMNy0KKojIQ2VzRE4LlLRyYg+xfLPH3hQOdFg5nR5bloYnHL6eTym1QgMbAZVhLg7pliREsgGzRkQutfJJT6RWFls+3SCCAQ+RUpxaUdvsIeGUBExTPm1onY4tW+5iG6x5odXPdxAVHk5Z3+1Z4dISuc6JPGsOokE9xDlqxlA2oNXIidp+BKPCk93sbiOmpJJxVi0i569ZJUc+c+LbcNcKMCPwKq2H6fcUdK0ApyTTTzw8lr/DWLZg0s1sid0ghkH8zO28mjgtZhF58Ti+QUvD0JdBtXZtdefR1MfbHi0feEL4JCkz42ESKybzJC8Z9xZOhPNhAZHLCgKoS/N2Ez+cEhWy1mHrzh8XXwdW7Mwf5FwFXqr7pssDtSyI2kZW48VZ4dISualr8Kg6E1OkYMBdtuPkCIvMizLg5Mm7FyHJGJE2JKc6kVfqI0+GS9QnN1E/MHImG60FQeS1IvJ+afvcqRJ5Rj9xrDJb1WADalLbD2HBzXzo6wVtd9SfZk4HEhB+K5GxVTtH/ovvfRJ/6+ceLo+mAxsONOjKMiVQ1YJVYUqQPzqARxW0OvnthwDrFFpsOBApTFqy8zX4cIdwWryOpl6EtrQekQfa7PcYbOdH5KIGFW5wQEldbGLU2z2EkHcQQh4lhHyIEPKdmzixMqiyBDtrxFIUQkOYm9OOlPyBFFlw/BAm/EqTuMdyH4pbTuSjWhH56sMlYt+GT2VY5oZNs4AVibw4tQJwIt9wjlwqiMipasEkHtyMiTV1YXMyykrDaZbIoy7vIGkwhYvleZ0CWy22TgJ9q3ZqZX/kglLgeFrcmy8FDhws+6EDAFQTRoWh5QE3zOrmGGYJ+JK5ZOnLrp1bKt4RropRKZGHaK2ZI4/4sOUJNWEY+feQYZiIKMnsoDsrbCIiDwH8PUrpSwC8CsB3EEJeuoHjlsIl+uwm5YiSfGj2gqCKCb1GBOY6U0iEQqrQi21r2zD9GhF5ifMhwIa7KhJZzW/FZ7lXs8SYayWIvvC6EXkwBcL8IQWnEZHPFIPLa4KqbAJO1uixunD9AGbOdl4U57PsIUjAUitlEbmj9mtH5IecwI/KiDzMETIh7SFeTOR0eoBjmu98KBDKJtRoOUfOrl1x0CEm3Jft7KY+7yNfR9UsOmSk4hqTpStwoNc249sk1iZySukNSumf8D+PATwK4K51j1sFHnRI4Xx0LUy08haEuHGrziBMpqJXUEf6xg66YbnybsQHEBcNlRAghKBjKMl7aiGwYcModVhcCUlE3q/+nkSmP8x9yakQeZGHh9aCRTy4VYcLF8Dls2CXHAQBmDxqy5q2TgIxeDk/Rw4AE6V+auXq0XvxkPrDOJoU1wHkyIGX4yckaRZM+JhW6Fo5op3EViAPWZ1CoqmgLBViGAamVAcpqbVMhLJzBXm+gBhXN5W7ha+zNEbkdc34NomN5sgJIfeBze/844x/ezMh5GFCyMO3b1efbVkETzIgLzzZHa6cy0uFUNVkxvYViTxIhulWeLK3drGFYamk/sQJIBGgnROBLaJrqitF5CR04EKDKp9CKWTV1ApQKtMfOpuV6CuRDQrCBEkLEOvEzTBxqgsxC1bOEPaYugIbRqb6Tw5tOFSHlVfs5BHuSO4B09tAiaoxjQfG78Ub5EfgDIsLnmxId3ZELuktWMSFXfKwk91jHKNTmjKMZBMadefUmY7PulayHoJpEEIwJm3IJVa2thfCoO5aEbnCO2S8AssBgI17s6me+ZA+K2zsDieEtAH8FwDfRSldsiejlD5EKX2QUvrg7u5mhgEHZFleHST+ItlETnhxq1SlJo7nCtVdeUQud6+hSxzcOiiOyk+4YZYkESCOgJKZkas6IEqBnRtlrY1Vi51AaS/5qpYEedAiF75kZgqXpMR8reYc1wyEgsgzyMNQ2GzHLHsIOXILi53Cg+YIfSAOSjt/0jB91t3hlgjV1MhFkGVrAYCoJiwSYFLStVLmfCggCsxeSvLvei40EkHKUbemMa3gST7xQuh0vRy5wR0QgyKfFcwGMF96iT4hRAUj8V+glP7qJo5ZBYFsLsmrRSpEySNyrQULLpyKRC4me1cR1Rh91qJ0dPvpwtfN+ay89XuBH7wH+Kk3AO/618Azf7LkcNc1lZXaD4u2y2tjnYi8KLVi1ZDpH38S+D9fzv6fgzCKodN8gVginc/yQKkJv+ChL0kELrRMy1wlYoXGrNY/gBX2O4aC25Rv8SumV+KYohuxh2Z4UhyRq3HBkG61hRYpCX7iCHpwgmN00Cmp/cSqhRbcuV1x4FQPmGy5C73Ek9zzfKgIVhoqISDU2qLomQcxgJlcZmUnYW7rPwXgUUrpj6x/StURyMZSrs1PXNxy5v7pFmRC4brVLroonlYR1XR2rgMARgfPFr5ujsj3/wJo7QGRD7zjnwM/8Rrghx4AfvXNwJ/9MhB6fNxb/YhczjFw2gj2XsqincF91d8jiPzmn+e+pJa688b7gePHgRsfyH2Jw4dK5AnEhA/NJlrHAj4sQc1p93SJAbJQ00EcQYk9BJIOOatjhGOrpeFmxNfgpJqfz8gNsMUFLXRa/B6NugjzbChUE0ZZ+6F9CAKKsTIo/D3Y8VowiYdp6sEgUph5AVgartKBUWJlG4td7hoReX/Ahi0LEVIeLFWBQzWQDfra18UmIvLPAfANAF5LCHk//++vbeC4pYik5Z5wMe1bzslpi6Jl5mShrM/w8r2iF9HbZTVe9+hG4evmiHx8E7j3VcCb3wn8/Y8BX/EQcP9rgI/+LvCrfwt470/w1Er9iFyN8j1n1sY9rwT+8Q2gvTxZPBe9e4F7XsUeWL/2bYC7vD2u47dydPNJAMDhradyXyN0AHkDqIWL5iaIXKyVPPMrnxhLQhixHS+Tpg8sDc/4fA1Oq9WYDqc+dgi7xrJd/J5Ch0+tBZOWNAhMWMTvagViIA6i8U6h1PGE++KSjW4GfLWbTLjPQyKXXyNH3tm5jljW8ZKXfFrh60xNhg19owNK6qJata0AlNI/AFDRNWmziBRzaQ6isCzVciJoYenqVSxuJYMASoQKAKD1WGolGN0sfN3ICXBXnx9vcgt4wevYn9u7wMu+mv0XR8APPg84eYp1rawQkauxi1A7pYh8FcgK8M2/Cbz7h4B3/2vgiT8AvuLHgfs+N3lJvwaRH956Clv8/3kxkxNE3AUv+4ZWk9Fj6xc7EyVtTkTuSwbMRcM2XiCLShSNA0vFUyP+gKhI5EcTD58OloIo0zfoKBjSrbKW3UnRGuS7hMAsjl4BVr+y4M11wYgIukpTQaj10KFjVvTNqHuEUcw6lRSs1bUCowvp7zwCqXOt8GWaIsGFATk8vZmqZbi0yk6ASYcNOt+TPLPjzCHygpmBWUhGXFV5srdYTo2UbGNHLp8O5E8BbwR0MuS/kgx0rgCTW+iaKmw/qq0+1GIX8WlF5KtCVoHXfC/wrb8DyBrwH94E/M4/SXrLRUQ+rJBaiccsCoxG+dfb9rn0O2eLnbhoZk3uSeOPfgz4k58rPh9hPJVD5IFkLE9b58rkvAeNwKCl4ZO2CYBUJvKT4wPohO3kDK+gAB8FUBHmOwWqFiRQhEVdGfycIrO8kUHS29BJADtlRSt2M1WUmL65CwM+qJedJ5/66w2VmEP/HhaAlCCQ9Y362tfFpSZyqjDFWbodK/LFzZS9INSaBjekwkDYBLKKsdSF4uQXoyils9TKmEfu7Rwfh/YVYHxrZU9y5sS34aESm8LdDwLf9vvAg38T+MO3AA+9Brj5wVqpFZk/MCU7n8gT07OcG1rjbn9lPcDH7/q/cetdP1H4mrhg0g/AivPqos8PL5DFJTu+gaXhwInYgIOKOXL3eJbiaxVFi/y882wtxNpPXCSzwM+JtsqJXNQlvOns4VlkNbwIrc9qUUc3s4vc9tyYt7MJZALJhLpBO+S6uNxErrUggQKpApJYEEt2nBxF9rdZSCrRFYsmtrYN08sncieIEER0nsizInKAETmPyAHUG/lGKQx4pd4V5wqtBbzp3wBf98ssovvZN6HLRYFViFx3b8/9PwtOIv3O/v6S6LloPVAKw9kHHZcQqF/80I9kA9qizw9/DylZX1stDbYfIW7tVo7I/RO2vkKioRsN86fq8HOgedEr/3nhw266Dw8qjHa/9LzErthP1SVojfts+9rzAADPPvV49qmkpwOtk1qpgVA2l3dbZ4hLTeTiS0/nN6lvI6YEppX9Ber854XRRQpJJboikQfGDnrxMNe7Y875cFKFyPdTVrY1IvLIh4x4rar9meGFbwBe+48B5xiyvY+OoVQi8nbA8r4tPz/adHwm/c4TiAnDo8VBB3PwxjDhoh8fFYtxkt1bnmGbtUzk4nNLUgpCFBQY25WJPOIPnmHr+dgmJ7ldJ8IjRCpIrQAoFLzQyT4OaC9pHy2CuOZBune/RkR+9733AwAOns2OyCdrzutcBZFsQqUBq22dAy41kQtHQjfVA0x85uJm5ogrFD1/slAWpNBBALVSngwA4tYedjHErVH203mOyEsj8j3AH6OvsPfUKnjWuDEuBDpsu4zRjWoy/ThCNx4CAHoFBGvz1EpeukNEbEU9wPYh64ox4CNy8oUoZWm4rJqOiIbLvHwGXGTjatWJnPDXTfovxA5OcDzJVswK7UVuflrMug1txHH2dXaHN3GbdnHPoEJqJMMtkBQYmy2iv3cPAGB6mK3XsL2QpVyBM4vIk0LxOYmCLjWRCzmvl3qyk9CGAyO/l5UvyqrTPJQC6XIW5M4V7JAT3DjJIXJ7gchlPd/TmxP8gLKIs466s/TmvGgQD7PxDSbTt0tk+vYhZMR4PL7C3C7tYebLXJ91reQSuaIhhFw4kHu0P2tvPN7PF3uxh76S/9BXTTbIN0p9jyKvrhW33Qm/lakyACbViFxxbiOChHDrARgkwPAku+ApRgLmPkz4DsOgPtwcxW04uoUD2sMDV8r1FrMxjSkiD2ukMPU2bGIhHmXrNSbp1MoZReRJoficZPqXm8gTefWMlKXQgUcKtndavQueNxU9D8bgGlrEw+3D7HavmRe5woi8czV/5iXv0RbqvDoRueiTzyWwi4Yuj8jH1SJy+4iN4HpceT4A4OQgm2Bd14FKosL+ZAf6cn93CtPDGZEP9/N71uWwWEmbWCunHxo8oMgTsAlstYTfygDwx5XmQ+reISZyD0qPtc/ZOfqGoMwYjgcDRZ7kkn2bE3l5H3hWy6dch8gB2PouNHsfYUYnl53uWjmjHWlC5EUpulPEpSbypGiSisil0IZbJEsX+cuKczvV2MmXLmegvcVumjx1p5i9meTI89IqAMuRA2j7rHhaJ0fu8msiV7DfvRCwtlk74ujZSkQ+5JHxuP+Sub8vQnQnFRG5R5YHHaThHz+T/Hl6mK/aZcZTRWtvOYgQ/uRlikbhX3JMuIFTBZl+KzjGVN2C2Wdr0j3J1jeIa5S7VkREDi97SlAcw/CHcPXtStbMIpii3nwAFoMASrXdb9y+ih0c44nD5QfwxEt1rZzRjjQpVp+TTP9yE7m+3EooRy6CQiLnMwMr5rLU2Mv16ciCPuA3zXF29LOUIy8kcvZvunsAiawWkcsVlHIXAoSwazG+gZ6plUfkh4xc1bteBgCY5BBsJGTzRUQuGVCigkLeyQ0ElPmg+MN81W6ZJYK42YNUbjgUQpiSKU6i2HlABZEXd9BQynxWPG0bFh9TFuT4rZQO6ebnbeXN7XSOICOC3L1SeE4JOLmmvdkV8RDM250uHqJ/HVdwjA/fXNaDiK4VKqlMt3AGSGyzm9RKfYjtaHqLpoQOvKKoSFYRQZrl5ApAKYVGnVx5dyZaLB0S5EQ/gqA6hgqMb+X3kAOsZ5jIkKb76BhqrT5y0dqlVrAWuDDoXJtLreS2ywHwOKHuvfDBub8vQrSZFtUKfLI8sSYNeXoTn8RV+FRGVKDa1WIXQY6nCzBLc3mp2Y6BM0VESeEEGmBmnJX4rZRE5GMvxDZOEJq7yS4xz29FdK3k7lo4kZvEz5TpR1yYZQ2KFZCz4/HvIpWGKDTtykB79x7skWN8+MZy8XnK6yJnlR8HUm6rTWqlPrSMOYhq7OTacQIACIFHjNmwgQJ4YQwTfj0i5+kQ4T2xiJEToGMoLCfonRRH5JLM1KKTm+iaSq1ip8h7VnFtvDDoXANGrNgZRLRwak88uokRNfHA/S+ERxXEOQSbdCcV3NSBXCzmMNx9HEk7OCIDyAXiI7XIrwSzh4mfIvLIm8CGAVMvjxwHloYbASfbElHQ0djDDjkBbe2CJIrj7LrNTA2dlyPnRA43c0rQ7RusbrC1d3fp75A+XrrArMb5NrpZUHrXoZMQTz37zNK/Tb0QXdkHOaOOFQCQ+XcbV2xr3jSeG0TupxdEvtOdQCAZUCpE5MJwKVfxlgVrGzEI5Bx1Z6LqLOshF+jMesnrpFbEw03LEUZdSHSvJxE5UCzTl+x9HGCAnqXhiPSTVrtFxBXaMDMVlym0/X2MtR2MlK1C8ZFOi4lcSaatz9IBkcvG8bUqTHEatDQ8WdE46/jkmPXPd/cARcMYLahO9ntEf7ieo71IInL4c46FAvs3GZFfuX5P6e+QPp4wmaKU8odgDV8gft8M959c+qepF6Ij+WcakcuJYvx8BjBfaiIX1rLpookW53tPCwSSCaWCCsvmhku1Kt+yAkcdoBUcZnqjLMnzy4icqzvrjnsLSwycLiQ6VwF/gm2FFaqK8uS6cxsjmQ3iHStb0N2cVIN4yBekViLZhE5zHuxxjH50BNe4AlffQSvIjmrjmEKnBcZTyK7pxP4UNq02V3XLUnHTIYDWLiXy6SFLNQkjt5E8gJ4jnKL+FDElMPI6Z2QVVFJhkey5naKwf8+995X+DgAASYZH9KRTxQvjQofKTHAjq/jkxtI5Tf0Ibck/Uw2FmqhVm9RKbRgt3o+a2qLp1C28mQDmY14UgQk4fgizwHAp9/jGNnZwgtvj5SHDJ07AKvtlPisCid9KvYhcbPEM6zIROWtB3OV980VE3goOYevMMtXRd2DlECwqeOWEismmyWRhehsyYkTtK/DNPfSj7F5sN2QP/aK1IvOHarCgRHago6WXC86u9gzcPHGB1k4pkYtiuzVg68tWB2gFeUTuFIroAACqCRMephldK97xDfhQ0OmVW9gK+JIBhUfkbhDBIH6pcdgceAC0R47xkVvzBc+pF6JFvDPrWAHS9bomIq8NU9fhUWWuUlxox8khZgaWwfFjmBUGwi4ibu1hN0cUNFolIp/eRs+QauXIRUrBbBUPjr1Q6LIoqx+x6LootdKLjuAbjDhCcxf9KJukSIlsHmDOg0bOegiGLAdLutdB2nvYJiOMp8t5UNtnAyxQsPY0frOnazrgSuQqA7Kv9UwcTHzE1m5pjjwYsRpNZ5s9HF19G51omPPiKWzoMIrOQWtxIl+OyOPJPsbyoHLHCcBNpnidihmblQdgc+ABUFbnSuK1coYRuVCrXurUCiHkpwkh+4SQD27ieFVhqAtzEOMIBvyZ8CIHscLMi/LkxgLMRc0vlU8vQu5cwQ5OWPS0gLkcuayXj0prXwFohCuKXc/90J/Cowoso7oq9dzBt8vdgEWbeQ8u6o3ZjS8Ky+09DDDC2F6+3onZf8HDmCpshmRWl8zoNsv/av27IHNhze2byz3rzJzLzzeewuxmj1OBBwlZaiVvXmca17mHvattlXatUE70uvDIN7bRpyeZvyMJHLhUzx01B7C5nS3Jn/MQB4AoptDdA/h6uQ95GqFsQOEFZptfu1o5bdUANbdwlzLEY4tELtwPzzBHbhgWYkoqezhtGpuKyP8DgDdu6FiVIUsEbnpWXuKgVvxkjxU2oSRPbizgcFVgnuFSHvT+NeySIW4Ml7/UEydAzxI95FfKo5gOI6ur0hBjL0RU8vBJEDhwoUFXLtGmixN522dEnpdaGR+wKFnmuxmldw0yoTjYX+5gqKIYpFoLFvHgh8sPSvuAEbm1fTesAYtuTzJmstq85a3IxVCkudI+PyRg8zqrFDuv91gxcKQMSvvIJTERqMV2LdTaxYBMMLaXawHM1kKHKhesFbWFjuQvze188sjGFk5A2/UGqodyCzqvUzliN1Mzgiada3i+PsZjN+d9yadexHZYZxiRm7oCB1plD6dNYyN3OaX03QDOZTyGS2YjlsS2pqztiKomDGT3xKYhBsLKNWXuev8qDBLgeDh/SdwgghfGs9RKWX4cSNoZd/nIrknFqJwENlwYIDW2u+cOzQKMHjT7FmSJYOhk+60cc5m8EF+ZfOj18a1lgk3aTIsexqowX1veFvvHzyCiBL29u9Dho/ymh8sPDMdzoZKoUEmr6yYiSuaEMHJow0a1Yuc1HpEfowfYh4VOe6p7iBHpJoIYwu0eRgfL/fbMWqBk56aaaEnBkiDoI7fG2CEnSeRfFZHCpg6FUQyXNxWQuhF05yquyyf48M3x3E5j6oVnTuQtTYENfa7x4ixxZuEaIeTNhJCHCSEP375dzfSnCnxiQOI3q5cYRRV/gVS1YJH5mYGZxxa92CU+GIsgnHydo3m14SjxWamg6hTgN+Agrue3UmpVcFHRuQ4yvoFugZWtUHG2dxixdnfvnvt5GmrkICAa68nPQdLfbS+rBOnoBg7Qw16vjcGVe9nrhss96x5/b9HuzdQV2DBm3ttgatCqqZVrPCLfj7sAjQE7P3Yy/EOMlVnaTu2yiHmS4bciR26xiA4ANHbPLEr0P3rzBNsYJYPHq4KqFlpwYQdRMvxDqk3k17AVH+LYDuYaC6ZeCI2ebbHT1GS4VAd9rrsfUkofopQ+SCl9cHe33jasCF6qJ1zcTKQkgiYqy4kWCU4AIBCKt7qiGr7NXJREL3mRVyJy9lDox+WdHGlIoVvLtfHCgMv0+5aGk5x2S/+YEfaA25kOuBAlS92pRE6p6ZmogWQN5JanN3GTbmG7rcHgkX88XiZy8dCXCx76pqjpZEjTSyfPg9WEtlsangnKe8nbwRFsdSv5uy78VjIeQkrowCclhUbVgkWW+8ifvnEDKomgdetF5JTfg7YXwfYCGCSo79TZuQrLP4SEOMmTxzGFHYRsyMMZRuQWH8DcSPRXhE+MZFZecjOVLAhJ54uoJCKPeJuYVlfmzsmXTrKJvK+EbIJ8FSLXWoDWQYe311WNyJWows15EdG9DoxuoGvmW9lGo1sIqIztXe5FIwh2Qd0Zx7SSrmDmorkckRvuPo6lLZY/VnSMSAdSRn5atJ0VuRgaqgyHarNhJWA7hjpCmGt9A096/DNy8uSUUvTiYdLVAwAt7rfinyy/R4ldhHJZasXKVHYeiXQW3zlWBt8V236YpETLjMOW0LkKQiNsY5TkyZ0ggk59ENAzLXZaGn9IVzTj2zQuPZGHsgE1ZtuqmdNd8RcoaS3oJJwb/pqFRLpcVx3J/VZU92CuM0YQ+TblvchVcuQA0N6D5TMir9q5okROoe/HhUXnGjC5hUFBu6U0vYVD0oehcVm71oINE9JChOqGrIhWJjQR9q1+RutYx9/HWJvtIMfKFgxvORIWQhC1gIxkicAlBogYTRj6kBHVMmW71jPxsSlf3zmdK7YfYRsniKwZkXd4GopmtC1qZbYWAKCyXvt0H3kYxXCO+cOzwqzO+Q9twYLLhorzAKz2zpcXx1/SniQR+dQLUxa2Z5taceisXnfW2FT74S8CeA+AFxFCniaEfOsmjlsFoWxCixci8hLHPznDxzwLkS8i/JpPdmsLMZGxRYc4nM6iShFN9yMuXqkSkfPXCWl41V5ytUIkeiHRvQbQCHdrUwxzflfNuY0Teb5tc6RsQVsgWGGxUDaAWphFRc4CkQcu2vEYnjlz9bO1nczRcpHLbmCtREnrkpT3OTdYqqNovN4z8NiER885veRHwxN0iJN4rABAp9OHR9XMdAyztSg5B60Fg3pzqZVPHtnoxyIoqeh8yCHpbVh8Vxzwa1f0EMwEJ/KX992kl3zqR2c+VAIANFmCQy45kVNKv5ZSeo1SqlJK76aU/tQmjlsFaXGPEFqUFSeVhMiLm/epJzoeai4ISUagby31kovpQN2wJpG396AKIq8YkWslvh8XFvzmvEs+zq0HWP4Bpuq8itDRlwnW5jd1mWJQWMiG3sJ6GLOce9SafU+htYsterRkvyAcOIVtRB4CkvL54fnUWKlOYNf7Jp5xdVBJyc2Rj3h7ppKylSWShGPSg5JhZaBTt/xhoppQ6Xxq5aO3Jtjh3VR1Uyuy3oJKIjiOjcgTEXnNFCa/f17SnuKj+xOEUYxp2ov8DHPkhBD4xIBcQTF+Grj0qZU4Ja8WzfhlUZHYwgVlRL7G3Mu4tct6yU9mOTNRvLNE5NipaPvZvgJpehuEVI/I9bimUu6igF+Ta9IxRk6QKdrqRcfwzfmtfGjuoh8fw0tpA0RbGy3RFagGI99ooXUsHjEiJ93U99S+gl0Msb8wk1UoacuCCF8yIAufH969UnZ+aVzrm6CQEBUMYbZ5ukMUOAVO5D50d8HKII6hwy9fK2oLCg3h+bPukI/y1kMqKfnjCnMgisKeM0nsJGpH5O09AATP00bwwxhPHE55aoVf3zMecxhIRqEd8mni8hO5arFBq5QmN2JZVCQ8uoMyg5ua46fSkPjszvQQ5hMnQEuTIU9vsWk4ZapOgfYVEG+EHS2qXOzU4ZUqXC8kujO/lZgyX+00ojDAgJ4gbi1s5dt72CVD7I9mRJOM/CrJlSYzJBfWw4SLgbTBXcnP1N41GCTA/sE8iYqHftloPT8lTU+8q2vkcoUoyNG2conc450pra15Ip8qA1jBgldMIHYFJWuFP2zk0E12Ix/Zn+B5+pSlcKR6VCIenoEzSUQ0tbtWZBVo7+EqYb/TYzfHmPqpeZ1nvP7DEhfN08SlJ3KqmJARA5GfSJ/LInKNK+wWI7BFVPHpyP2M3lXsLfitzDkftgtmdS6C5x/v0yfVHBAphUndlc773NHaBYiMAa8jLO5Ajm8/C4lQyJ15Ild6V9EjNm4dzQYNOAErdpbpCkQxmy4M5Lb5HFBrZ2bPanB152hhtBz1q6mKI9mAJiJyvl7rWEAIUdBEGeQSuRj0IHxWBFxtG51F069EDV3eRw4gaRkEWER+jzapX+jELI0SOOOZ6d0qxNu5il54CFki+PDNMSZelBrzdtZEbkCt4OF0Grj0RJ7M//OnSYRj5vkqi7cI8yK/mMglschX2KKR9h52yAluDtOplYCJgSY3E+l9JXAiv0cbVYrIo8CFTOiZRyQbgSQD7SvohSyXu5gnF8OPtcE8SZlc5Xmcks+zYqdbGukZLb6DW5juEgyfgUtVDLZmRNXj4qMldacgo5LPihSTiVVSn1dH0Xilo0MiwBHpAZMcf3FO8OZgfo353G8Fab8V8TuXrfFkShAreIZRjE/cnmJPGtVvPcRs1xx6k2QA9UrE27kGeXoTn7LTwmM3x7DPqWsFACLFgkb9QsXtaeHSE3kyB9GZMEtQqpVOWyHJNI8SIo9sxJBYGqQu2legIcRoOCsujQSRj2/Wq/Jz0r9LGWNcgcgdIWw54xzhxtC9hrbHOjIWHRAnvJDXWog2u7y9bpoaep0oBkvSHYZusJmc/nx+k46exU26hb3uLFptcwVjcDIvPkp6w0sjchM6nY/Ic6fXZ0CRJVzpGrgddVlEnmGCJdsHGMMCWTgXau1ARYjYGc5+xs9h8bVLSM3tnHohnji04UcxetFx0m5bB8lQGGdS2SMpE52rwPgmXnS1wyNyZj3NPuRsA5lkAM05DGC+/ESuzYomJGC+FYZS4lshiNwvvuBy6MIj1QfCzkHM7hzNREEjN5VaqVroBBLSvyqfVEqtuFMmjqjdNnlR0LkGw2VEvhiRC5/t/sJYMZNH6G5K3Wl7PgwSlBKlIkvcRXP+wS5Pb+EWBtjrzIhc2C9Eqe8VYOZXbAp8cYoiViyoCIEoSFI5db18rvUMPBt2WA3HXy7Y6+4BTqSM+guPnMXQCWDWuVWmhp5NCfIw9SN89NYYAIXhHyVK5jog/DuJfXvWsrdKBN25Bkxv46W7Jp48snF77J1bRJ4UjM9Bpn/piTwZaGtPIAUOXOiQyuTO/Mm/eOMuQonL5d254DcNHd9KDH1OnADbegy4w3qpFWsbIBL2yLBSaiXx/ajbBXBR0LkG1WZEuUjkIVdvDhaInPDrSVPy+aoCMQBwiLHUA2y6t3BItuYNrcwBAqhLszvl0IFP9NKHftKhEthJ/3Tu0OMcXOubKXXncnrF9I8wScnzBdQOW5OT4+VrVOrwqaVSK16Ij9yaoEemkGJ/pYhcPBhib7p+RA7gU/tsl/OnTw7ZmLfUOZ8VkhReScr2NHDpiVy4zfnuBKSKixuQmuJdHJGrUfFU9EJwIu9Gx0kUfeIEuEvhlpt1InJJBlp72KbDSu2HnnBtrEkQFwbda5C8ExjwlhwQpcktjNCCulggTAYMzwg25JL7KkTpwpgJdQCAUrT9A0y0BZIiJFPdKUcu/ComZaJu4dvwXXZ+Ws0pTnf1TXzM5qSXkSfvREfwtGUi17iVgXOcjsjZOZTuCjjJmvAZke+P8Wk9/t2skCNPBjD7k1lf/UrFTnYfvdhiO4sPPD1ET/GBCrujjUM8OJrUSn0kcxCdCeTIZqmQMiQRefEWSI3r+WDMIbGfHeLGyEEQxbD9KGmVqizPT463h358hLEXlg7EEDdnXdfGCwM+8u1u5WQpIlfd2xjKyyQFWcVE7s3N7hQinbIuJoCZr8lpnwx3CI16cM3ltAETHx3OWacqkQ2/yu4tudntmYBthdTKjZAXaDMi8n48RGAsD3oQY9/8lJnbbNdSRuTs3y2wYudHb43x6X2ewliha2UWTNmQIr6bqdnCCCCJyHdxhJYmwwtjdOSAPRTO2MI58aJvUiv1IdqYQneSOMmVIumJzb/glFJo1EMor7g9M/qgRGGdKyduQki7GLJ/r6rqFGhfQTc8AqXAJGOSeRpC6FSFwC4kuADnfn20tANpeQeYqtnTaByNDUcWDzphsVBlMEjafA1AMoovai3vnAJzFzsYzj1kmCVC+doT2+/QnSB0+RQns16wcK1n4oD22F8WjLMcx0GfTBFnkGt36yqbYjNO7VoqEzm7Zwzi4cQO8PjBFC/q8KLtOhF56ECtet9mgUfk0uQWXniVPdw6Us1pQxuC2NXQjLrFaePSE7mYXh16NpTIrZbTJgQeMSAJ86IMeCGb17myOlKSmLoT80Q+iGvK8wU6VyobZ4WeUBnWNPu6KOA35/O0k6WulW50BM/IjgBDaxfbOMHBlEWKQjFYpXvHl02o8YzI6Yh1v0jdZSInnavYJUPcTIm91Nir5G0jPLc9Z4LIm8KpOFQijet9A0fgs1gXjLOGt1lXj5RBroOOiWO0QexZFC+0FGpZ54w2i8g/9OwIQURxn8FzwavkyBUTMQikwIYSe6vXoqwdQFKA8Q28mBN5W/LPpfV25qLZROS1oRkzcY8au+Uubhy+tBCBLUCoAukaMneJ95LfSBF5LzoEJBUwM9IDRWhfgeEfgiAuzZPPfD8ua2qFkefd8nzUG4QRtugQkZVN5LS1h10ME38b6lUXmoSSwTysOdwj1o+uplSdAmrvKrbJGLeOZ7a3zNumnIyklM8P9djQ41aFoRJpXO+bCKDAU7pLxllicISSMbGnrSs4QheyMyP/mcNnCZGncuTvf2oIALgmjwEiA1bNtQwAkgSf6FAiG9o6KUxJYmnK8U28+Cp7uLXI2Q6VEBC7miw75NPGpSdyPaXS1OIK5j8cgWRAifMjcpsPcKVrPNlJ5wquySPcPHET8m0Hhyx/Xjcf2L4CiUbYwriUyMXNmQhdLhuMLqC1cVWaN866fXgIi3iZUTLA5PO7KREWDaoLTQLZgp6KyO1DRuStnWUiN7n0fchnhEYxhU6r7d4UXaQCp4gDGw7VYdWMyLdbGjRFylR32ryQKYZgpEEIwYk0mPNbEWpovSwNpxgACHqKj4/dnoAQYAtDNhO0YPpSEQLJhBzZlXczueDDSF7EI3Jmy3D2Ebkoqpdaf5wCngNEzsnKn0Kj1a1bw4Wt9CLcYLWBsHNo7WFXOsHN0Swit9zb9dMqwNzszjIHRCHyuLREDgCdq9ijR3OplWRWZz+byI2tazCJj8NjRlTEr96fHCkpoQ6A4PgZHNM2dvr9pde2d7i6k4/yE8IjWuZXglnUFrgTEJ/pHuoSOSEE13oGjkh/icjFVKr2VvbotYkygBnMXCKpP4VL1Zm3e/6HAqqFjhyAUuDeLQuKfbBaWkWcq2xADhxY66QwgUQUJFIrJvHOJUeumbP6x1nj0hO5aTBVHvVtGLS64x/zvPBy/51NRffrD4RNo72HQTzEraGdRNGau78mkZe3IAqhiVl3IMZFAp/HmP5dJ8L7ZDubpFpcpj/h6s6kK6nCdxgrFhvYy0FHN3CTDrDXWW5nVfmOQKg7xRT4Ki6GSU3HZZYSDnS09HqpFYB1rhzEnSUij3khs7ubfY0cbYB2OJz9wK8+/BmalfRoP7DXYYXWFcRAAoFsQY1dmMQtN+0qQudaMh7wFfcN0JX9MxcDAbPmgqAh8voQI5YQTGHArxQVASwCM+HCD+PMfxc58toDYdNo70FGhOnoIInI5emtFYmcRT57qCAKCmz4VIGslERZFxnd6+iHBxh7zNcDmKk6hd/JIiQuChKzO2eKwSpEbsKAD8Tss5TpTezTeVVnAv45MVd3Oj43aqrwOcL7PPImkEIHDq1f7ASA6z0TzwSd5eES09uYUh2dTi/zfYGxgxadAoGw0rVZwVWtcA6qiRYn8hdeabMe9jUi8kixYMGFCb+Wle8SOleZyC5w8Mvf9lewrYbnEpHr/CGdFNnPEJuaEPRGQsiHCSEfI4R8zyaOWRW6IsGGDs1j/dlVc9qU37hOztxOx/Ohk6BculwETr6Gd4gbJy46agTiHNfvIQcWIvLi1AoJbDhV+ukvMjrX0PJvA6BJKik8EUR+T/Z7Evk8ax1M+sIrFL7oQg+w4d7GPtlC18yIljl5CXWnHYSVd29i+x17U0ghT61UIdEFXO+beMpvMQILZ6IpxbmNIemD5PRQRyYfyGGzgicJ+MOkEpG30CacyPfaa0fksWLCIt76OW0hruODQBDY55Ijt3QFNtVLPZxOA2sTOSFEBvCjAL4IwEsBfC0h5KXrHrfG58NLEXlVmS9VLVjw4ATZRC48KJQaFqNL4Df8DjnBR26N8Xydb7lWicj1NqC1cV0ZlRpnSYEDDxUUrhcZnWuQaYgtjJPdDJnsw4cCycrxcedELoYjy5GDqKrpmSp6gKdAFKIVHGKq7WYToqJhKvdgeIwMbY95YFfx0zZNCxEloL4NJXTgEQOKXP82vNY3cEB5C6I9K14a3iFGcj//jXxNihQMCR040KArFc5BNWFyIn/RFoDQXSsipwq7Bw3irWe5LO4nYc/g2+fStWJpMmzosyL7GWITEfkrAXyMUvoJSqkP4JcAfNkGjlsZLjFgcMP8yhG0asHkU7yz4CdCiTXyzCKKxhAfvjnG8zTelrQKkfPjXZNPSlMrUuTAW7Uv96KA56GvkiMMbUYeqrOPoTTIV+yZA0REhu4dgFLKBlBLZjWFn7B6cCbAdB8SYrhmPkm5+ja64RHcIEpmv8oVHvqmpsCGwYg8rqh7yMD1HFFQKziCnSHPF5C77HcS3S1yaMMlRrk/EQCoFusaIsDzTZ4+WEUMxEG1Fky4sFDtIZiLpYh8em4RuQtt5k1/htgEkd8F4KnU35/mP5sDIeTNhJCHCSEP376d7aO8KnyiJ5NPKhcnNYuZ5OekVsIahku54NtO0Wlyj8qHHqxB5Huk3AFRCSvKxS8yuEz/Cpm1IJreYa6qEwAgSXC0bfSjY4zckE2Hr9jFRDSW3/TsUeaszkWE1h72yDH2Rx48MfS7ApEbqgwXGhDYUKPVB2TPReSpgmc3GsLVd3LeBeg9QeQsepUiF0HVNJxmoacE+LwHdmctjKvI8wVUCy3iwYRf6drlIh2Rhz4Qn0+O3FJl2FQvtf44DWyCyLMe5UtmIJTShyilD1JKH9zdXePLz4AvGWhHQwDV5NjsdRZM+LmpFdFCpJaMjSuE0QeVNezyAbXXZTGodlUi38NOhWLnOpHehQG/OdNE3o0O4eaoOgVCk81KvXniQqsomwfmhTrgszqlbnbnBwA+u5O1lvp8rVQx5zI1cbNPoa4xIPt638QBeEQujLOiED06QmjmE7kYwCH8VmrJ41UTu3qEn/2WVwIT7teyRkRO9Bb6GEMidD2nTnMAyDp7AK8wPm9TMHlqhYSXMyJ/GkC6+nQ3gGdzXnsqCGQTChghKxUXhKS3YBEPtpdNionibR3jKUJAWnu4zh0Pr5BjJie2CqLKInSuYhAflRL5OpHehUHnKigIrpIjnDgBHF+oOkuIo30Fu+QETx7ZMOEhquiVIyfma2MEQ7Z8tQxVp4Da4zL9EwchT61UWXumym521RtCAkW0Ygqga6hwRQqFR+TeeB8SoYVRcr/Xh031xG+l1kNfbc2c/cQuYI0cuay3oBF238rrpFYIYam48c1kWMe5dK0oErPRvqQR+fsAPEAI+RRCiAbgawD8xgaOWxlhaiFWNemXk4EU2YWJyBPTW9Z8srf3cF1hufFteryaqjN1LIva8OziYgqTPF9yIpdVoLWLKzjGiR1gfzjGDhmBlKSllO5V7JATPH4wgYnqSl85pcpzDp9CSCV0tvM/y9y6CwYJcHx4kOzetAq7N1ki8IgO3WepwHWEMN3uAD7Rkhz5iPfPSwXTpwaWhgPaTcbBqZFTedcC1Zx5bU/2AZDVgxIAUsrfRVnX4K3DiTzxNj97IieEwCcG5ALrj9PC2kROKQ0B/M8AfhvAowD+M6X0Q+setw7SN2vVBTHzRchu3k8G8a47wLi9l6RWBvHR6vlxICmeam5xjUGrKBe/6CDda7hLZqmVo30eJWdIz9MwB9ewgxM8cXsMi1T3ypkJdSbwh89iH33sdvMf4gZPUUyPnkGYyNyrPfQ9YsAS6so1COfawMKQ9BLjrAmf/KNl+KwIDFoaDtGDzNsPNepVf+hrVioi32ckLtcXMwmkU1HaugETl+nPUivnMx3Lly4pkQMApfS/U0pfSCm9n1L6A5s4Zh2kF6JacZjCTCqdHd3SGvLuQrT3MKAs+uqGh6vnx4HkvdaCF/Yi6ihcLzQ613BVGmLoBBhxVWdrKz/dAQBy9woUEmP/1o1aXjkzoc4UGD2LfTrAboaqU0CMfPNPbiQCkKr+7z4xYHF1JV1jfd3VN3A77iWiIHfICpjWVv4aa2kyjtCD5h0ClEKnLsKqa0W12Hi5OGZ5+TpzZ7MOlwq61JrDNZbQOf/UCsBsB9Qo38PptHDplZ3A/Pa0alSkpnzMM5EQ+ZqE2NpDOzyBhBgtb0WfFQFeWNqix7ndNpRSGGuafV0YdK5hDyxH7h6ziLybo+pMIOZSHj3L/E8q3tBCXh17UyjTm3zockEvPv8e6ehWsnur2kIXyAYUyuockrb6+rrWM3Ez6iDmaZKQK0072/nTpwghGCsDmP4hEHqQQBFXjcjFmgqdtcVAwDyRKzUGUGeic5XNLxVF2HModgJAKBV7OJ0WnhNEniYtraK/iG4Ig5vswgSpoQosRPsKJETYwzH0YLiR1AprZ8wueHoBl4tf1sHLaXSvo09HmNrTRNXZyvFZScCvkWzfhkn8yuQq8tuxN2GqTvSx3Sog8vZM3RnXfOinrZalNQjsWs/AAe0l4h463odHFQwGxQTrqAO0wmGS76780E+NqcNkf61CJzCTtAM12obzIHrJjz7O/n9O6z9STGg038PptPDcIPJURG5UjMjFZPUoZ5pH0gu6bmTLo5aXSp9kf1+HyFs7oJCwR45ze8ltx4ZMKMg5RSQbBb9WyvQWIJSIZdv5lAjLhFe5HdU0TfhUhuwcwojGGKt7kItEMkYfIdFgerdnBcCKn5XuKKp6flm43jdxiC4k+wCgFJJ9G4fooWsVe+x4+g5kxMCI2fBW3nWmxtRhenut1kMAkNNiu3XvM3FfHX58M8dbEbFsQqde4tlzVnhOELm4gVyqwtIryLGB5IumOQY3kojI1x3gyonlhz5X/H0NIpdkBMY2dpEfkTsT1uoorSOwuCjgoiDD2Yfq7GNEuoBS8v0KczEyhAW3cn+yqTLzNWvCHrhegaoTAEAIXH0bWxjCsXkwUJEQ06lAeY3hH9f7TN0p0RBwjqG6BxhK+T4ryedbvM98yH7X6hE5P+/pASPzdcRAwHzUvG4KU0TkgsjPQaIPYNZOesa95M8JIiectCrbcQLJIkqKmgtQIuaDsXKroADffm6NHmN/XyciBxBZu4VWth4v3q4lsLgo4DL9ln8bhneAiVphEo3WRiibuIscQCa0cvuooUqwYaAzZeRWpOoUCK097GEIe8qtFyoS4lxxfo3viaVWZiPfTP8IYznHhyaNFifyY/a7Vk5riF3e8ePs/2tG5HN57HWJV9xXR+cbkScPpBxeOS08J4hcbE8d6NXMf4DkgtOcAcxy5CCQNmA8JRb7zT9j/1+TyGnnKvbIMHdup2czUqkqjLrQ4FHWVnSATngIV68QARKCwNzF8wgrelVRWwKAoTDFZccRszpLcvFAIj6ivo0QCut9r4A4RTLrKIcNVYZv8D7u6T7a4RFsrbyvW+6wNRkdPQGgRnpHkBSP5NfNkW80Itc7gNaeCZXOKSJPHiBnbJz1nCByIe5xoZduKxPwaIDkPDnrzP8shN5h6Znhk3y+Yb58ugqkzpXCYqcviLwigV1omANEkoYr5Bg7VVSdAu093FuTyCWJwCU6JK4Q1kraHAFA7V/FHjmGCQ8+qfHQV9NdVut9T8mQ5ckt9OLhjNiLPr7LWycPWGRdefcmyPH4Cfb/NbtWoKV+901E0CJIkvWVx8+ti2R+QdCkVmpD5sZWfh0Pbn4zOdMxDibLVWYtrqF4KwIhs6h8HVUnh9q9ih2cYGRnV8ZFX3zVfvoLDULgmldwlRxhjwyTgQ5lUHtXcTfhgpcaROnx9TOlOvr98jSOMbgLW2SCLrHh13nop0hr3QHZxoCT18FHISNGbJaTq9XfY1a6PLVSWb0sHkDHG4rI+XWIiVp5N1MIkSc/x44t0SWVCArPCM8JIhc9qF4tIud59ekYn/+v3oF//duP4YTPh6SUQo1rKN7KIBb8mmkVAJB716CSCOHkMPPfg0Qu/hxIrQAIrat4gDwDnQRQS1SdAkr3KlTh4VEjxeQR9n3fogPsdsu/e5k/WO4mtxHINSJyfrPbVIelr0dg3cEeIhDEN7mYukLeetA2cIQutAkzLZWr+gmJB5CIyFvr7S6h6ACRNleYF0R+jh1bUuLZc7bj3p4TRC6EBbWsW2UFkDV88yuv4LUv3sOPvuPj+Nx/9Xa85W0fxcHEh0U2KHMXLXMbIHJxo1IhfFhAxCNy/TLP60whbl/FA4TP6qyQ7gAwpzis43Mt1s8tulWo6kzAv8/nkVu1TMrE9psNXl5d4g4AVwdtHNEOwpsfBAAo3XIi3+J+KwpXIFYuuAoiP3kaMLfWj6IJYaS7qcKkuL/OMSIXg2i8hsjrI5leXTcVoloYKAH+3dd9Jn7rOz8Pr3r+Nn74dz+C1/7wO2GsO9k7DZFL3AiRc8FLaphAGhHP+Rvr2O9eIEi961AI68nt7lQk8nRbXA0iF97lt9DPHLq8BP5QvUqOa+3eRATqUB2Wvl4u91rPwCHtQR2yfLdRYdcyaKmzoRSYV1gWQhAkjdbvWEmOuUkiFxH5ORI5743P83A6LTwniFwsxKpDBBJorWRG40uudfET3/gg/ut3fA4+494BLHibKxiKCHGdHvKFY+UZZ8UeW0D6ut4VFwRpK1mjwFZ2DmnRUI2bWkTVN6tG5Knvs85DX4jRbOhorRmR39U3cUC7IHwEgDXIl+cLDCwNh+gmf6+chktfy3V7yJMPtzYfkZ8rkc9cNM8SzwkiN3gaIa4dkZtL/Z4vv6ePn/uWV+L5fQlXtiv05FZBa/MRueEeZP6z6IuvU+S7yNC3Ut4qVaPANJHX2GaLqPpE2YFRZRhxiszqELloDXWgVdc95OBa38QhHzARUgm97fJrZGkyjkkfAOBTGYZe8b6RZNYRAqxtmJVAbW0uFXIBip1JdiDPw+mU8JwgclM3EVAZkVLzC1St3DYhJXTW938Q2GSOXG/DIwZaQTaRJ7/PefXRbhgSFwX50ACjV/JqjjTh1yh8ifXjl6k6BRQNntrnf65B5Dxqc6gBa00iv9LRccRFQYfoYqtdfh6EENhcXOXWEdEBs86VTaVWencD3Yo7rTJcgIhcBFCh1xB5bZiagu8LvxUPb72p3htViwkIogxxTeBsbkHc+9nAi98E3P2KjRxurO6gEx5l/hsJpvCgnlsf7cbBo6yptl1tiDKwQOTVCVbIq6MaKbCQdyTRGg9OlUdtLtGhyuvdgooswdFZ7/gB7aFvVitA+jojchs6zCq7DwHxe24qtfLXfwL40rds5lhJsfP8ghjDEi6aGfqUOAIe/U02V3TDWGsVEUL+B0LIhwghMSHkwU2dVF2Ymoxfjl6Ncef+em+86zOBZx4GfvSVwJ/+AhBxkQ2lLHe+KSJv7wJf8wuAVUFiXgGOzgYMZ3mSk9CBhw0oUi8KOJGX2temoeiA0Wd/rpMjV1lkS3r3lLxyBrXLyKPTrl5cXqnLqgCUi8xOpAGkIqOvFAI+19OhWj0i33RErneAdS1sBVSTRfebOrcVYOkqHKrNE7k7At7zo8D/9XLgP/2PwGO/ufHPXTci/yCArwTw7g2cy8pQZQJZIvUWJAC84QeAr/55llP79b8NvOWzgEd+FvBGAOiFtYL1jV3s4hhve3Q/6X0XkEMb3mUfvJyGagDmVtKzXRntKyzdUUOA9eGdN+Cb/H8Ebas6kWt9JuXf2apeTzF0ExElm1EOg6l9AWCiVj8HyiNqB0bN1Aq/J9YVA50WvuWtwOf/g3P7eDGTFf4UOHoceOv3Aj/yUuC3vw/o3g38jf8IvPTLNv65a5XMKaWPAqguiz8lEELw/V/yUrzivpoRryQBL/kSlvb4yG8D7/qXwH/7u8A7/g/27xd0OIM+uI7u/h/idT/3MADg/t0WPuPeAT7j3j7u9ux6CtfLgC/4R8D2C+q9p70H2Dl1hBzIZhvvil+Gz6vSsZL+HKDWWrF0FSO04CqbaRE1envAs4BbwWdFQObn7UCrVtgVEL/nuvL800L/3nP9eEtjLpr3PPtbwFt+GSAS8Je+Evjsvw1c/4xT+9z1ep9qgBDyZgBvBoB77938xf7Gz75v9TcTArzojcALvxD4+NuAd/0rYHJzVgW/YLj77ucBH7Hxe1/wOH6ffBb+v1sK3v7YPn7lkafxk+oEofkcI/JXfVv99/TuYaO/asBQGKFVaj0USAps1aNrU5XxZv9/QXvrfnxVnRPMgbXFdgWhWV1pqXHhkIuaeXrtgkfk5wxLU/BRuosdPAt87ncDr/ifEhfP00QpkRNCfg9AVvXnH1NKf73qB1FKHwLwEAA8+OCD+QMnzxOEAC94PXD/65gM+Zyf7rm4/zXAwz+DF/zxP8YLAPzNq58O+lfegJtXPh/qOxS0tM3k4i81Xv/9gDOs9RaRYtjr1HgQio6kGhG5oUl4H30xXmVu5nvqXX0+/vfg66Fd+aLK7+l2OhhTs36ePkmtXNCI/JxhqBK+KfgefPtfuR/f9bpPO7PPLSVySunrz+JELhQIAbY+5bzPIh93fRbw3R8E9v8C+OjvAB/5HZA/+De4Rn+I/fv9rzvf87sI6Fyt3e4paiyFszoXsUJqRXzOumIggesDEz8V/TV8Z7+C9S7HoKXhCXoFR3LNh4lqsUJy2YCPOxSEECiqgXF4tl1jZ5ZaabBhEAJc+Uvsv8/9bsA5Bj7+duBjbwde8NrzPrtLiS944S6+4VXPw/O2atRGdl4EaB1g54HKbxE56XXFQAL3blnQFAnP265+3gNLwzf534O9dg9/vc6HffpXA1c/tfY53kkwNSV3OPppYS0iJ4R8BYC3ANgF8P8SQt5PKf3CjZxZg3owB8Cn/nX2X4OVcN9OC//7l9ckqc4V4PuervUWVZagymRjEXnf0vCuf/DqWimhrZaGI3Sxp9Vs/Xvg9ey/BrmwNBkfeGqId354H599/zZ05fSj83W7Vn4NwK9t6FwaNLhjcFffxF2DDZmyAbjWq3esQYulRmp1rDSohC952TX81B88jm/+mffB0mR83gM7eN1LmMvqTvt0NB5NaqVBg3PAf/s7n3smkVoeBhZTgNbWXjQoxT/4whfj77z2Abzn44f4vUdv4W2P7uO3P3QLhDAvp3/yxS/FZz1vQz5OHA2RN2hwDugYG5iIswZMVYauSBvL0zeYh6HKeM2L9/CaF+/hn385xYeeHeFtj+7jbY/dQsfYPO02RN6gwR0IQgi2WhoM9Tlht3ShQQjBp97Vw6fe1cN3vr56UbwOGiJv0OAOxd97w4twvf8cE4/doWiIvEGDOxRf9Vk1jMgaXGg0+6oGDRo0uORoiLxBgwYNLjkaIm/QoEGDS46GyBs0aNDgkqMh8gYNGjS45GiIvEGDBg0uORoib9CgQYNLjobIGzRo0OCSg2RNYj/1DyXkNoBPrvj2HQD1hjE+N9Fchxmaa8HQXAeG5/J1eB6ldGk807kQ+ToghDxMKX3wvM/jvNFchxmaa8HQXAeGO/E6NKmVBg0aNLjkaIi8QYMGDS45LiORP3TeJ3BB0FyHGZprwdBcB4Y77jpcuhx5gwYNGjSYx2WMyBs0aNCgQQoNkTdo0KDBJcelInJCyBsJIR8mhHyMEPI9530+ZwVCyE8TQvYJIR9M/WyLEPK7hJCP8v9vdprrBQQh5B5CyDsIIY8SQj5ECPlO/vM76loQQgxCyHsJIR/g1+Gf8Z/fUddBgBAiE0L+lBDym/zvd9x1uDRETgiRAfwogC8C8FIAX0sIeen5ntWZ4T8AeOPCz74HwNsopQ8AeBv/+3MdIYC/Ryl9CYBXAfgOvgbutGvhAXgtpfRlAF4O4I2EkFfhzrsOAt8J4NHU3++463BpiBzAKwF8jFL6CUqpD+CXAHzZOZ/TmYBS+m4ARws//jIAP8v//LMAvvwsz+k8QCm9QSn9E/7nMdjNexfusGtBGSb8ryr/j+IOuw4AQAi5G8AXA/jJ1I/vuOtwmYj8LgBPpf7+NP/ZnYorlNIbACM4AHvnfD5nCkLIfQA+A8Af4w68Fjyd8H4A+wB+l1J6R14HAP8WwD8EEKd+dsddh8tE5CTjZ03v5B0IQkgbwH8B8F2U0tF5n895gFIaUUpfDuBuAK8khHzqOZ/SmYMQ8iYA+5TSR877XM4bl4nInwZwT+rvdwN49pzO5SLgFiHkGgDw/++f8/mcCQghKhiJ/wKl9Ff5j+/IawEAlNIhgHeC1VDutOvwOQC+lBDyBFiq9bWEkJ/HnXcdLhWRvw/AA4SQTyGEaAC+BsBvnPM5nSd+A8A38T9/E4BfP8dzORMQQgiAnwLwKKX0R1L/dEddC0LILiGkz/9sAng9gMdwh10HSun3UkrvppTeB8YHb6eUfj3usOsAXDJlJyHkr4HlxGQAP00p/YHzPaOzASHkFwG8Gsye8xaA7wfwXwH8ZwD3AngSwP9AKV0siD6nQAj5XAC/D+DPMcuJfh9YnvyOuRaEkE8HK+LJYMHYf6aU/m+EkG3cQdchDULIqwH8fUrpm+7E63CpiLxBgwYNGizjMqVWGjRo0KBBBhoib9CgQYNLjobIGzRo0OCSoyHyBg0aNLjkaIi8QYMGDS45GiJv0KBBg0uOhsgbNGjQ4JLj/wdPOTKjuH+s5AAAAABJRU5ErkJggg=="
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  }
 ],
 "metadata": {
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3.8.2 64-bit"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  },
  "interpreter": {
   "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}