Experiment.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2 as cv\n",
    "import numpy as np\n",
    "import Functions as f\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "img = cv.imread(\"Puzle_color.jpg\", cv.IMREAD_GRAYSCALE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Puzzles of the same shape are assigned to the same class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "y0 = np.array([2,4,1,2,5,2,5,6,6,7,5,8,6,1,9,5,6,3,8,8])\n",
    "y45 = np.array([2,1,5,4,2,8,2,6,6,6,1,5,5,8,9,6,7,5,8,3]) \n",
    "y90 = np.array([2,5,8,6,1,8,6,1,2,4,8,9,5,6,2,3,5,6,5,7])\n",
    "y180 = np.array([8,3,8,6,5,9,1,6,8,5,7,6,6,2,5,5,2,1,2,4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "target_list1 = [y0, y45, y180, y90]\n",
    "angle_list1 = [\n",
    "    np.array([45, 180, 90]),\n",
    "    np.array([0.86, 1, 0.84]),\n",
    "    np.array([1, 2, 3])\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train1, y_train1, m1 = f.augmentation_img(img, \"MONO_ROT\", target_list1, angle_list1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Reading second foto to increace X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "img2 = cv.imread(\"Pm1_3.jpg\", cv.IMREAD_GRAYSCALE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "angle_list2 = [np.array([])]\n",
    "y0_2 = np.array([5,6,6,5,6,5,5,6,6,5])\n",
    "target_list2 = [y0_2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train2, y_train2, m2 = f.augmentation_img(img2, \"MY_BLUE3\",target_list2, angle_list2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Merging 2 X_trains "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, y_train = f.Merge_train(X_train1, y_train1, X_train2, y_train2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. Creating X_test from image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_img = cv.imread(\"Red_3.jpg\", cv.IMREAD_GRAYSCALE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = f.create_X_test(test_img, \"RED\", X_train.shape[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. Validating to find best parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test_red3 = np.array([1,2,3,4,5,6,5,7,5,5,6,8])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "neighbours_array = np.arange(1,10)\n",
    "accuracy_list1 = []\n",
    "accuracy_list2 = []\n",
    "accuracy_list3 = []\n",
    "accuracy_list4 = []\n",
    "y_test_here = y_test_red3\n",
    "X_test_here = X_test\n",
    "\n",
    "for k in neighbours_array:\n",
    "    clf1 = KNeighborsClassifier(n_neighbors=k, weights='uniform', metric='euclidean', n_jobs=-1)\n",
    "    clf2 = KNeighborsClassifier(n_neighbors=k, weights='distance', metric='euclidean', n_jobs=-1)\n",
    "    clf3 = KNeighborsClassifier(n_neighbors=k, weights='uniform', metric='cosine', n_jobs=-1)\n",
    "    clf4 = KNeighborsClassifier(n_neighbors=k, weights='distance', metric='cosine', n_jobs=-1)\n",
    "    \n",
    "    clf1.fit(X_train, y_train)\n",
    "    clf2.fit(X_train, y_train)\n",
    "    clf3.fit(X_train, y_train)\n",
    "    clf4.fit(X_train, y_train)\n",
    "    \n",
    "    pred1 = clf1.predict(X_test_here)\n",
    "    pred2 = clf2.predict(X_test_here)\n",
    "    pred3 = clf3.predict(X_test_here)\n",
    "    pred4 = clf4.predict(X_test_here)\n",
    "    \n",
    "    accuracy_list1.append(np.sum(pred1 == y_test_here) / pred1.size)\n",
    "    accuracy_list2.append(np.sum(pred2 == y_test_here) / pred2.size)\n",
    "    accuracy_list3.append(np.sum(pred3 == y_test_here) / pred3.size)\n",
    "    accuracy_list4.append(np.sum(pred4 == y_test_here) / pred4.size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 5))\n",
    "plt.plot(neighbours_array, accuracy_list1, color=\"g\", label='uniform Euclid', marker='o')\n",
    "plt.plot(neighbours_array, accuracy_list2, color=\"r\", label='weightned Euclid')\n",
    "plt.plot(neighbours_array, accuracy_list3, color=\"b\", label='uniform Cosin', marker='x')\n",
    "plt.plot(neighbours_array, accuracy_list4, color=\"y\", label='weightned Cosin')\n",
    "plt.grid()\n",
    "plt.legend()\n",
    "\n",
    "plt.title(\"Accuracy for Red_3\")\n",
    "plt.xlabel(\"k_neighbours\", fontsize=13)\n",
    "plt.ylabel(\"Accuracy\", fontsize=12);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "6. Predicting codes for puzzles and view it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = KNeighborsClassifier(n_neighbors=3, weights='distance', metric='cosine', n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='cosine',\n",
       "                     metric_params=None, n_jobs=-1, n_neighbors=3, p=2,\n",
       "                     weights='distance')"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "7. Download the predicted image or view it right here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "f.view_code(test_img, \"RED\", True, clf, X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv.imwrite(\"experiment_red3.jpg\", f.view_code(test_img, \"RED\", True, clf, X_test))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}