Presentation2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "13c32e98",
   "metadata": {},
   "source": [
    "# Distributed Algorithms: || k-means\n",
    "## <span style=\"color:green\">2nd notebook</span>\n",
    "### <span style=\"color:blue\">Group 20: Dall'Ara, Verguizas I Moliner</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55cd704e",
   "metadata": {},
   "source": [
    "###  <span style=\"color:red\">KDD Dataset</span>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "431fbb24",
   "metadata": {},
   "source": [
    "The dataset we have worked with is the kdd99 dataset which contains network traffic data of different types of malicius activity attacks. Therefore, our goal is to group the points of our dataset in k different clusters based on their similarity. Following we can see the schema of this dataset as well as the different types of attacks labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6ce2110c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "root\n",
      " |-- duration: long (nullable = true)\n",
      " |-- protocol_type: binary (nullable = true)\n",
      " |-- service: binary (nullable = true)\n",
      " |-- flag: binary (nullable = true)\n",
      " |-- src_bytes: long (nullable = true)\n",
      " |-- dst_bytes: long (nullable = true)\n",
      " |-- land: long (nullable = true)\n",
      " |-- wrong_fragment: long (nullable = true)\n",
      " |-- urgent: long (nullable = true)\n",
      " |-- hot: long (nullable = true)\n",
      " |-- num_failed_logins: long (nullable = true)\n",
      " |-- logged_in: long (nullable = true)\n",
      " |-- num_compromised: long (nullable = true)\n",
      " |-- root_shell: long (nullable = true)\n",
      " |-- su_attempted: long (nullable = true)\n",
      " |-- num_root: long (nullable = true)\n",
      " |-- num_file_creations: long (nullable = true)\n",
      " |-- num_shells: long (nullable = true)\n",
      " |-- num_access_files: long (nullable = true)\n",
      " |-- num_outbound_cmds: long (nullable = true)\n",
      " |-- is_host_login: long (nullable = true)\n",
      " |-- is_guest_login: long (nullable = true)\n",
      " |-- count: long (nullable = true)\n",
      " |-- srv_count: long (nullable = true)\n",
      " |-- serror_rate: double (nullable = true)\n",
      " |-- srv_serror_rate: double (nullable = true)\n",
      " |-- rerror_rate: double (nullable = true)\n",
      " |-- srv_rerror_rate: double (nullable = true)\n",
      " |-- same_srv_rate: double (nullable = true)\n",
      " |-- diff_srv_rate: double (nullable = true)\n",
      " |-- srv_diff_host_rate: double (nullable = true)\n",
      " |-- dst_host_count: long (nullable = true)\n",
      " |-- dst_host_srv_count: long (nullable = true)\n",
      " |-- dst_host_same_srv_rate: double (nullable = true)\n",
      " |-- dst_host_diff_srv_rate: double (nullable = true)\n",
      " |-- dst_host_same_src_port_rate: double (nullable = true)\n",
      " |-- dst_host_srv_diff_host_rate: double (nullable = true)\n",
      " |-- dst_host_serror_rate: double (nullable = true)\n",
      " |-- dst_host_srv_serror_rate: double (nullable = true)\n",
      " |-- dst_host_rerror_rate: double (nullable = true)\n",
      " |-- dst_host_srv_rerror_rate: double (nullable = true)\n",
      " |-- labels: binary (nullable = true)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.datasets import fetch_kddcup99\n",
    "\n",
    "kdd_data = fetch_kddcup99(as_frame=True)[\"frame\"]\n",
    "kdd = spark.createDataFrame(kdd_data)\n",
    "kdd.printSchema()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "694416bb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique labels: ['smurf', 'pod', 'guess_passwd', 'teardrop', 'normal', 'land', 'ftp_write', 'back', 'nmap', 'ipsweep', 'multihop', 'neptune', 'warezmaster', 'imap', 'satan', 'phf', 'portsweep', 'loadmodule', 'buffer_overflow', 'perl', 'rootkit', 'warezclient', 'spy']\n",
      "\n",
      "Total amount of labels: 23\n"
     ]
    }
   ],
   "source": [
    "# separate labels\n",
    "labels = kdd.select(\"labels\")\n",
    "\n",
    "unique_labels = labels.distinct().rdd.flatMap(lambda x: x).collect()\n",
    "\n",
    "unique_labels = [item.decode('utf-8').rstrip('.') for item in unique_labels]\n",
    "\n",
    "print(f\"Unique labels: {unique_labels}\\n\\nTotal amount of labels: {len(unique_labels)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9270cfea",
   "metadata": {},
   "source": [
    "###  <span style=\"color:red\">Data Preprocessing</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3dcc9689",
   "metadata": {},
   "source": [
    "Before applying our kmeans algorithm to the kdd99 dataset, we need to apply some preprocessing steps. Our processing has consisted on removing the cathegorical columns and the columns with 0 variance as well as to standarize the remaining columns in order to bring all features to a similar scale."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1bdc4f2b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                                                \r"
     ]
    }
   ],
   "source": [
    "#drop cathegorical columns.\n",
    "categorical_columns = ['protocol_type', 'service', 'flag', 'labels']\n",
    "kdd = kdd.drop(*categorical_columns)\n",
    "\n",
    "kdd.createOrReplaceTempView(\"kdd_table\")\n",
    "\n",
    "# drop columns with variance equal to 0.\n",
    "for coll in kdd.columns:\n",
    "    if spark.sql(\"SELECT VARIANCE(\" + coll + \") FROM kdd_table\").collect()[0][0] == 0:\n",
    "        kdd = kdd.drop(coll)\n",
    "\n",
    "#standarize the remaining data columns.\n",
    "for col_name in kdd.columns:\n",
    "    avg_val = kdd.agg({col_name: 'avg'}).collect()[0][0]\n",
    "    std_val = kdd.agg({col_name: 'stddev'}).collect()[0][0]\n",
    "    kdd = kdd.withColumn(col_name, (kdd[col_name] - avg_val) / (std_val))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3148756c",
   "metadata": {},
   "source": [
    "###  <span style=\"color:red\">Distributed Analysis: Random Initialization</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f665a7e4-1cef-4074-9fba-d2746d551c0d",
   "metadata": {},
   "source": [
    "Now we implement a parallelized version of LLoyd iterations together with a random initialization of the clusters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ea892268-be83-4d56-9c04-09b93fbee262",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to calculate Euclidean distance between two points x and y\n",
    "def euclidean_distance(x, y):\n",
    "    return np.linalg.norm(np.array(x) - np.array(y))\n",
    "\n",
    "# Function to find the index of the minimum value in a list x\n",
    "def argmin(x):\n",
    "    return x.index(min(x))\n",
    "\n",
    "# Function to initialize K-Means with random centers\n",
    "def random_choice_init(data, k):\n",
    "    # Select relevant columns and take k random samples\n",
    "    three_dim_array = np.array(data.select(array([c for c in data.columns if c not in ignored_cols])).rdd.takeSample(False, k))\n",
    "    final = three_dim_array.reshape(three_dim_array.shape[0], -1)\n",
    "    return final\n",
    "\n",
    "# Main K-Means function\n",
    "def kmeans(data, centers, max_iter=100, perc=0.999999, return_cost=False):\n",
    "    count = 0\n",
    "\n",
    "    # Get column names\n",
    "    cols = data.columns\n",
    "\n",
    "    # Initialize cost history \n",
    "    cost = [float(\"inf\")]\n",
    "\n",
    "    # Main K-Means loop\n",
    "    while count <= max_iter:\n",
    "        print(\"Iteration\", count)\n",
    "\n",
    "        # Define a UDF to find the index of the nearest center for each row\n",
    "        argmin_distance_udf = udf(lambda row: (argmin([euclidean_distance(row, c) for c in centers])), IntegerType())\n",
    "\n",
    "        # Calculate the \"centr\" column based on the nearest center\n",
    "        data = data.withColumn(\"centr\", argmin_distance_udf(array([c for c in cols if c not in ignored_cols])))\n",
    "\n",
    "        # Update the centers based on the assigned clusters\n",
    "        \n",
    "        centers = np.array(data.groupBy(\"centr\").mean().select(*[col(\"avg(\"+c+\")\") for c in data.columns if c not in ignored_cols+['avg(centr)']]).collect())\n",
    "\n",
    "        # Calculate the cost value per element, then sum\n",
    "        get_cost = udf(lambda row: str(min([euclidean_distance(row, cen)**2 for cen in centers])))\n",
    "\n",
    "        data = data.withColumn(\"cost\", get_cost(array([c for c in cols if c not in ignored_cols])).cast(\"double\"))\n",
    "\n",
    "        # Calculate the new cost and add it to the cost history\n",
    "        finalcost = data.select(\"cost\").groupBy().sum().collect()[0][0]\n",
    "        cost.append(finalcost)\n",
    "        \n",
    "        print(\"Cost function values\", cost[count])\n",
    "\n",
    "        # Increment the iteration count\n",
    "        count += 1\n",
    "\n",
    "        # Break if the cost change ratio is below the specified threshold\n",
    "        if cost[-1] / cost[-2] > perc:\n",
    "            break\n",
    "\n",
    "    # Return the result based on whether cost history is requested\n",
    "    if return_cost:\n",
    "        return data, centers, cost[1:]\n",
    "    else:\n",
    "        return data, centers"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "689a931f",
   "metadata": {},
   "source": [
    "We will compute both the cost function and the time it takes for each iteration. We will then average by a number of iterations(in our case 5). We will repeat this for every configuration of partitions and cores per worker that we choose. In our case these values will be [1, 2, 3 and 4] possible cores per worker and [8, 16, 32, 64] number of partitions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "bf780f52",
   "metadata": {},
   "outputs": [],
   "source": [
    "def avg_cost_time_results(k, data, partitions, iterations, func, L=None):\n",
    "    \n",
    "    total_cost_history = np.zeros(len(partitions))\n",
    "    total_time = np.zeros(len(partitions))\n",
    "    \n",
    "    for part in partitions:\n",
    "        part_cost = np.zeros(iterations)\n",
    "        part_time = np.zeros(iterations)\n",
    "        for i in range(iterations):\n",
    "            start_time = time.time()\n",
    "            data = data.repartition(int(part))\n",
    "            data = data.persist()\n",
    "            r_centr = func(data, k, L)\n",
    "            result, final_centroids, cost_history = kmeans(data, r_centr, max_iter=15, return_cost=True)\n",
    "            part_cost[i] = cost_history[-1]\n",
    "            elapsed_time = time.time() - start_time\n",
    "            part_time[i] = elapsed_time\n",
    "#            total_cost_history[np.where(partition==part), i] = cost_history[-1]\n",
    "#            total_time[np.where(partition==part), i] = elapsed_time\n",
    "        total_cost_history[np.where(partitions==part)] = np.mean(part_cost)\n",
    "        total_time[np.where(partitions==part)] = np.mean(part_time)\n",
    "    \n",
    "    return total_cost_history, total_time\n"
   ]
  },
  {
   "attachments": {
    "a4e60f96-0f06-47b7-8e42-c1ad0b546b9f.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "a2a51fba-3029-4fd3-a2ac-9d67e525c3c6",
   "metadata": {},
   "source": [
    "![randkmeans.png](attachment:a4e60f96-0f06-47b7-8e42-c1ad0b546b9f.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ae17f1f",
   "metadata": {},
   "source": [
    "###  <span style=\"color:red\">Distributed analysis: Parallel Initialization</span>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68643e07-0130-499c-8d03-1053ac4ec5dc",
   "metadata": {},
   "source": [
    "Now we do the same but with a kmeans parallel initialization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2b7435b0-1b51-4162-a58e-69f62c05991e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def kmeans_Lloyd(X, centroids, k, max_iters=100, tol=1e-7):\n",
    "    \n",
    "    \"\"\" Run the sequential LLoyd algorithm given data X, centroids and k number of clusters.\"\"\"\n",
    "\n",
    "    costs = []\n",
    "\n",
    "    for _ in range(max_iters):\n",
    "        # Step 1: Assign each data point to the nearest centroid\n",
    "        distances = euclidean_distances(X, centroids)\n",
    "        labels = np.argmin(distances, axis=1)\n",
    "        \n",
    "        # Step 2: Update centroids based on the assigned points\n",
    "        new_centroids = np.array([X[labels == i].mean(axis=0) for i in range(k)])\n",
    "        \n",
    "        # Calculate the cost function (sum of squared distances to centroids)\n",
    "        cost = np.sum(np.min(distances, axis=1))\n",
    "        costs.append(cost)\n",
    "\n",
    "        \n",
    "        # Check for convergence\n",
    "        if np.linalg.norm(new_centroids - centroids) <= tol:\n",
    "            break\n",
    "        \n",
    "        centroids = new_centroids\n",
    "    \n",
    "    return centroids, labels, costs\n",
    "\n",
    "\n",
    "\n",
    "def kmeans_ll(data, k, L=1): \n",
    "    \n",
    "    \"\"\" || kmeans initialization of k centroids elements. X is the data, L the oversampling factor.\"\"\"\n",
    "    \n",
    "    \n",
    "    cols = data.columns\n",
    "    n_samples, n_features = data.count(), len(data.columns)\n",
    "    centroids = np.zeros((1, n_features))\n",
    "    \n",
    "    # Step 1: Choose one random point as the first centroid\n",
    "    three_dim_array = np.array(data.select(array([c for c in data.columns if c not in ignored_cols])).rdd.takeSample(False, 1))\n",
    "    centroids[0] = three_dim_array.reshape(three_dim_array.shape[0], -1)\n",
    "    \n",
    "    # Define a UDF to find psi\n",
    "    # Calculate the distance using the UDF and collect the result\n",
    "    distance_udf = udf(lambda row: ((float(euclidean_distance(row, centroids[0])))), DoubleType())\n",
    "\n",
    "    distance = data.withColumn(\"distance\", distance_udf(array([c for c in cols if c not in ignored_cols])))\n",
    "\n",
    "    # Sum the distances\n",
    "    psi = distance.select(spark_sum(col(\"distance\"))).collect()[0][0]\n",
    "    #print(psi)\n",
    "\n",
    "    \n",
    "\n",
    "    # Calculate psi\n",
    "    #psi = spark_sum(distance_udf(array([c for c in cols if c not in ignored_cols])).collect())\n",
    "\n",
    "    \n",
    "    for j in range(1, int(np.log(psi))):\n",
    "\n",
    "        # Step 2: Compute distances from each point to the nearest centroid and define probabilities\n",
    "        min_distance_udf = udf(lambda row: (float(min([euclidean_distance(row, c) for c in centroids]))), DoubleType())\n",
    "        data = data.withColumn(\"probs\", min_distance_udf(array([c for c in cols if c not in ignored_cols])))\n",
    "        # Calculate the sum of the \"probs\" column\n",
    "        total_sum = data.select(spark_sum(col(\"probs\"))).collect()[0][0]\n",
    "        # Normalize the \"probs\" column by dividing each row's value by the total sum\n",
    "        data = data.withColumn(\"probs\", col(\"probs\")*L / total_sum)\n",
    "        \n",
    "        # Sample and add centroids\n",
    "        # Generate a random column for filtering\n",
    "        data = data.withColumn(\"rand_col\", rand())\n",
    "\n",
    "        # Filter and select the rows where 'rand_col' is less than 'probs' to get n_samples\n",
    "        selected_data = data.filter(col(\"rand_col\") < col(\"probs\")).select([col(c) for c in data.columns if c not in ['rand_col','centr','cost','probs']]).collect()\n",
    "\n",
    "        #print(\"centroids\", centroids)\n",
    "        #print(\"selected_data\", selected_data)\n",
    "        if selected_data != []:\n",
    "            centroids = np.concatenate((centroids, selected_data), axis=0)\n",
    "                \n",
    "    if len(centroids) > k:\n",
    "                \n",
    "        #A SORT OF K MEANS ++ INITIALIAZATION FOR Lloyd ITERATIONS ON CENTROIDS       \n",
    "        cc = np.copy(centroids)\n",
    "        p_centroids = np.zeros((k, n_features))\n",
    "\n",
    "        for j in range(0, k):\n",
    "            \n",
    "            # Define weights\n",
    "            w = np.zeros(len(cc))\n",
    "\n",
    "            # Create new column with the index of the nearest center for each row\n",
    "            argmin_distance_udf = udf(lambda row: (argmin([euclidean_distance(row, c) for c in cc])), IntegerType())\n",
    "            data = data.withColumn(\"index\", argmin_distance_udf(array([c for c in cols if c not in ignored_cols])))\n",
    "\n",
    "            #w = np.array(data.groupBy(\"index\").count().collect())\n",
    "            # Group by 'centroid_inddex' and count the occurrences, renaming the count column\n",
    "            centroid_counts = data.groupBy(\"index\").count()\n",
    "            # Order by ascending index value\n",
    "            centroid_counts = centroid_counts.sort(\"index\")\n",
    "            # Calculate the sum of counts\n",
    "            total_count = centroid_counts.select(spark_sum(col(\"count\"))).collect()[0][0]\n",
    "            # Normalize the \"probs\" column by dividing each row's value by the total sum\n",
    "            renormalized_counts = centroid_counts.withColumn(\"count\", col(\"count\")/ total_count).select(col(\"count\")).collect()\n",
    "            probs = np.array(renormalized_counts).flatten()\n",
    "\n",
    "            next_centroid_idx = np.random.choice(len(probs), p=probs)\n",
    "            p_centroids[j] = cc[next_centroid_idx]\n",
    "            cc = np.delete(cc, next_centroid_idx, axis=0)\n",
    "\n",
    "        #print(len(centroids), len(p_centroids))\n",
    "        # K MEANS Lloyd ITERATIONS ON CENTROIDS\n",
    "\n",
    "        final_centroids, __, _ = kmeans_Lloyd(centroids, p_centroids, k)\n",
    "    else:\n",
    "        final_centroids = centroids\n",
    "        \n",
    "    return final_centroids\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48b3f649",
   "metadata": {},
   "source": [
    "Again we compute the cost and time and we average it for 5 iterations. We then repeat the process for the same configurations of partitions and cores that we worked with random kmeans initialization ([1, 2, 3 and 4] possible cores per worker and [8, 16, 32, 64] number of partitions)."
   ]
  },
  {
   "attachments": {
    "90d1c442-97f5-46cb-94b2-656406071371.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "e4bbe7e3-192b-4d01-83da-3935182a18a3",
   "metadata": {},
   "source": [
    "![parallelkmeans.png](attachment:90d1c442-97f5-46cb-94b2-656406071371.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ecad17d-380b-4046-a929-964afbbf3824",
   "metadata": {},
   "source": [
    "###  <span style=\"color:red\">Comparison between Random and Parallel initialization K-means algorithm</span>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f1f3c8a",
   "metadata": {},
   "source": [
    "Now we can plot the average cost for both random initialization and parallel initialization of the kmeans parallel implementation for the two best configurations. As we can see, the average cost is lower when we apply a parallel initialization at the expenses of a bigger execution time."
   ]
  },
  {
   "attachments": {
    "3af579e0-0109-427a-8575-dcc8dbf407c1.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "f8709f44-2018-487a-8d49-f247c603d8a9",
   "metadata": {},
   "source": [
    "![comparison.png](attachment:3af579e0-0109-427a-8575-dcc8dbf407c1.png)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}