zeroR.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Baseline Modelling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ZeroR Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# load raw data files\n",
    "import pandas as pd\n",
    "tf_df  = pd.read_csv(\"train_features.tsv\", sep = '\\t')\n",
    "tl_df = pd.read_csv(\"train_labels.tsv\",  sep = '\\t')\n",
    "vf_df = pd.read_csv(\"valid_features.tsv\",  sep = '\\t')\n",
    "vl_df = pd.read_csv(\"valid_labels.tsv\", sep = '\\t')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a dictionary to count the number of genres\n",
    "genres = { 'Action': 0, 'Adventure': 0, 'Animation': 0, 'Children' :0, 'Comedy': 0,  \n",
    "           'Crime': 0, 'Documentary': 0, 'Drama': 0, 'Fantasy': 0, 'Film_Noir': 0, \n",
    "           'Horror': 0, 'Musical': 0, 'Mystery': 0,'Romance': 0, 'Sci_Fi': 0, \n",
    "           'Thriller': 0, 'War': 0, 'Western': 0\n",
    "        }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# count the number of genres in training data set\n",
    "for i in range(len(tl_df.genres)):\n",
    "    genre = tl_df.genres[i]\n",
    "    genres[genre] += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to get evaluation metrics\n",
    "from sklearn.metrics import classification_report\n",
    "def getMetrics(trueLabel, predict):\n",
    "    print(classification_report(trueLabel, predict,zero_division = 0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Action': 86,\n",
       " 'Adventure': 104,\n",
       " 'Animation': 30,\n",
       " 'Children': 106,\n",
       " 'Comedy': 583,\n",
       " 'Crime': 237,\n",
       " 'Documentary': 207,\n",
       " 'Drama': 713,\n",
       " 'Fantasy': 298,\n",
       " 'Film_Noir': 78,\n",
       " 'Horror': 244,\n",
       " 'Musical': 154,\n",
       " 'Mystery': 270,\n",
       " 'Romance': 791,\n",
       " 'Sci_Fi': 417,\n",
       " 'Thriller': 598,\n",
       " 'War': 241,\n",
       " 'Western': 83}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# class distribution in training data set\n",
    "genres"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy :  0.1705685618729097\n"
     ]
    }
   ],
   "source": [
    "# calculate accuracy based on the algorithm of ZeroR (the most observed genre is 'Romance')\n",
    "predict = []\n",
    "count = 0\n",
    "total = len(vl_df.genres)\n",
    "for i in range(total):\n",
    "    if vl_df.genres[i] == 'Romance':\n",
    "        count += 1\n",
    "    predict.append('Romance')\n",
    "print(\"Accuracy : \", count/total)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation of ZeroR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "      Action       0.00      0.00      0.00         6\n",
      "   Adventure       0.00      0.00      0.00         2\n",
      "   Animation       0.00      0.00      0.00         3\n",
      "    Children       0.00      0.00      0.00         3\n",
      "      Comedy       0.00      0.00      0.00        38\n",
      "       Crime       0.00      0.00      0.00         5\n",
      " Documentary       0.00      0.00      0.00        18\n",
      "       Drama       0.00      0.00      0.00        43\n",
      "     Fantasy       0.00      0.00      0.00        18\n",
      "   Film_Noir       0.00      0.00      0.00         4\n",
      "      Horror       0.00      0.00      0.00         8\n",
      "     Musical       0.00      0.00      0.00        10\n",
      "     Mystery       0.00      0.00      0.00        18\n",
      "     Romance       0.17      1.00      0.29        51\n",
      "      Sci_Fi       0.00      0.00      0.00        16\n",
      "    Thriller       0.00      0.00      0.00        28\n",
      "         War       0.00      0.00      0.00        21\n",
      "     Western       0.00      0.00      0.00         7\n",
      "\n",
      "    accuracy                           0.17       299\n",
      "   macro avg       0.01      0.06      0.02       299\n",
      "weighted avg       0.03      0.17      0.05       299\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# evaluation \n",
    "getMetrics(vl_df.genres, predict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Genre distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization of genre distribution in training data\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig = plt.figure()\n",
    "ax = fig.add_axes([0,0,1,1])\n",
    "genre = genres.keys()\n",
    "train = genres.values()\n",
    "ax.bar(genre, train)\n",
    "ax.set_xticklabels(genre, rotation=45, ha=\"right\")\n",
    "plt.legend([\"Train Set\"])\n",
    "plt.xlabel(\"Genres\")\n",
    "plt.ylabel(\"The number of instances\")\n",
    "plt.show()\n"
   ]
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}