evaluation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Precision(eval,emo,epsilon ):\n",
    "    \n",
    "    total = 0\n",
    "    print('===============Precision=============')\n",
    "    for i in emo :\n",
    "        retrieved = len(eval[eval['dominant']== i])\n",
    "        relevant = len(eval[(eval['dominant']==i)&(eval['real']==i)])\n",
    "        total += relevant/(retrieved+epsilon)\n",
    "        print(i,'retrieved: ', retrieved, 'relevant:', relevant,\"precision:\", round(relevant/(retrieved+epsilon),2))\n",
    "        \n",
    "    \n",
    "    print(\"\\nAverage Precision: \", round(total/8,2))\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Recall(eval,emo,epsilon ):\n",
    "    total = 0\n",
    "    print('\\n===============Recall=============')\n",
    "    for i in emo :\n",
    "        relevant = len(eval[eval['real']== i])\n",
    "        retrieved = len(eval[(eval['dominant']==i)&(eval['real']==i)])\n",
    "        total += retrieved/(relevant+epsilon)\n",
    "        print(i,'relevant: ', relevant, 'retrieved:', retrieved,\"Recall:\", round(retrieved/(relevant+epsilon),2))\n",
    "        \n",
    "    \n",
    "    print(\"\\nAverage Recall: \", round(total/8,2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def MAP(eval,emo):\n",
    "    print('===============MAP=============')\n",
    "    print(emo)\n",
    "    retrieved = eval[eval['dominant']== emo]\n",
    "    retrieved_rank = retrieved.sort_values(by = [emo],axis=0,ascending = False)\n",
    "    count = 0\n",
    "    total = 0\n",
    "    min = len(retrieved_rank) if 10>len(retrieved_rank) else 10\n",
    "    iter = 1\n",
    "    for i,row in retrieved_rank.iterrows():\n",
    "        if row['real']==emo:\n",
    "            count+=1\n",
    "            total += count/iter  \n",
    "            mark = 'O'\n",
    "        else:\n",
    "            mark = 'X'\n",
    "  \n",
    "        print(iter,\" - \",mark, 'precision :', round(count/iter,2),'(', count,'/',iter,')')\n",
    "        iter+=1\n",
    "    \n",
    "    \n",
    "    print(\"\\nMAP : \",round(total/count,2))\n",
    "        \n",
    "                     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "===============Precision=============\n",
      "Anger retrieved:  7 relevant: 1 precision: 0.14\n",
      "Anticipation retrieved:  20 relevant: 14 precision: 0.7\n",
      "Disgust retrieved:  8 relevant: 3 precision: 0.37\n",
      "Fear retrieved:  11 relevant: 3 precision: 0.27\n",
      "Joy retrieved:  15 relevant: 11 precision: 0.73\n",
      "Sadness retrieved:  7 relevant: 4 precision: 0.57\n",
      "Surprise retrieved:  2 relevant: 0 precision: 0.0\n",
      "Trust retrieved:  6 relevant: 1 precision: 0.17\n",
      "Love retrieved:  20 relevant: 10 precision: 0.5\n",
      "\n",
      "Average Precision:  0.43\n",
      "\n",
      "===============Recall=============\n",
      "Anger relevant:  3 retrieved: 1 Recall: 0.33\n",
      "Anticipation relevant:  18 retrieved: 14 Recall: 0.78\n",
      "Disgust relevant:  4 retrieved: 3 Recall: 0.75\n",
      "Fear relevant:  3 retrieved: 3 Recall: 1.0\n",
      "Joy relevant:  26 retrieved: 11 Recall: 0.42\n",
      "Sadness relevant:  15 retrieved: 4 Recall: 0.27\n",
      "Surprise relevant:  0 retrieved: 0 Recall: 0.0\n",
      "Trust relevant:  15 retrieved: 1 Recall: 0.07\n",
      "Love relevant:  10 retrieved: 10 Recall: 1.0\n",
      "\n",
      "Average Recall:  0.58\n",
      "===============MAP=============\n",
      "Anticipation\n",
      "1  -  O precision : 1.0 ( 1 / 1 )\n",
      "2  -  X precision : 0.5 ( 1 / 2 )\n",
      "3  -  X precision : 0.33 ( 1 / 3 )\n",
      "4  -  X precision : 0.25 ( 1 / 4 )\n",
      "5  -  O precision : 0.4 ( 2 / 5 )\n",
      "6  -  X precision : 0.33 ( 2 / 6 )\n",
      "7  -  O precision : 0.43 ( 3 / 7 )\n",
      "8  -  O precision : 0.5 ( 4 / 8 )\n",
      "9  -  O precision : 0.56 ( 5 / 9 )\n",
      "10  -  O precision : 0.6 ( 6 / 10 )\n",
      "11  -  O precision : 0.64 ( 7 / 11 )\n",
      "12  -  O precision : 0.67 ( 8 / 12 )\n",
      "13  -  O precision : 0.69 ( 9 / 13 )\n",
      "14  -  O precision : 0.71 ( 10 / 14 )\n",
      "15  -  O precision : 0.73 ( 11 / 15 )\n",
      "16  -  O precision : 0.75 ( 12 / 16 )\n",
      "17  -  O precision : 0.76 ( 13 / 17 )\n",
      "18  -  X precision : 0.72 ( 13 / 18 )\n",
      "19  -  X precision : 0.68 ( 13 / 19 )\n",
      "20  -  O precision : 0.7 ( 14 / 20 )\n",
      "\n",
      "MAP :  0.65\n",
      "===============MAP=============\n",
      "Joy\n",
      "1  -  O precision : 1.0 ( 1 / 1 )\n",
      "2  -  O precision : 1.0 ( 2 / 2 )\n",
      "3  -  X precision : 0.67 ( 2 / 3 )\n",
      "4  -  X precision : 0.5 ( 2 / 4 )\n",
      "5  -  X precision : 0.4 ( 2 / 5 )\n",
      "6  -  O precision : 0.5 ( 3 / 6 )\n",
      "7  -  O precision : 0.57 ( 4 / 7 )\n",
      "8  -  X precision : 0.5 ( 4 / 8 )\n",
      "9  -  O precision : 0.56 ( 5 / 9 )\n",
      "10  -  O precision : 0.6 ( 6 / 10 )\n",
      "11  -  O precision : 0.64 ( 7 / 11 )\n",
      "12  -  O precision : 0.67 ( 8 / 12 )\n",
      "13  -  O precision : 0.69 ( 9 / 13 )\n",
      "14  -  O precision : 0.71 ( 10 / 14 )\n",
      "15  -  O precision : 0.73 ( 11 / 15 )\n",
      "\n",
      "MAP :  0.7\n",
      "===============MAP=============\n",
      "Sadness\n",
      "1  -  X precision : 0.0 ( 0 / 1 )\n",
      "2  -  X precision : 0.0 ( 0 / 2 )\n",
      "3  -  O precision : 0.33 ( 1 / 3 )\n",
      "4  -  O precision : 0.5 ( 2 / 4 )\n",
      "5  -  O precision : 0.6 ( 3 / 5 )\n",
      "6  -  X precision : 0.5 ( 3 / 6 )\n",
      "7  -  O precision : 0.57 ( 4 / 7 )\n",
      "\n",
      "MAP :  0.5\n",
      "===============MAP=============\n",
      "Love\n",
      "1  -  O precision : 1.0 ( 1 / 1 )\n",
      "2  -  O precision : 1.0 ( 2 / 2 )\n",
      "3  -  X precision : 0.67 ( 2 / 3 )\n",
      "4  -  X precision : 0.5 ( 2 / 4 )\n",
      "5  -  O precision : 0.6 ( 3 / 5 )\n",
      "6  -  X precision : 0.5 ( 3 / 6 )\n",
      "7  -  X precision : 0.43 ( 3 / 7 )\n",
      "8  -  O precision : 0.5 ( 4 / 8 )\n",
      "9  -  O precision : 0.56 ( 5 / 9 )\n",
      "10  -  X precision : 0.5 ( 5 / 10 )\n",
      "11  -  O precision : 0.55 ( 6 / 11 )\n",
      "12  -  O precision : 0.58 ( 7 / 12 )\n",
      "13  -  O precision : 0.62 ( 8 / 13 )\n",
      "14  -  X precision : 0.57 ( 8 / 14 )\n",
      "15  -  O precision : 0.6 ( 9 / 15 )\n",
      "16  -  X precision : 0.56 ( 9 / 16 )\n",
      "17  -  X precision : 0.53 ( 9 / 17 )\n",
      "18  -  X precision : 0.5 ( 9 / 18 )\n",
      "19  -  X precision : 0.47 ( 9 / 19 )\n",
      "20  -  O precision : 0.5 ( 10 / 20 )\n",
      "\n",
      "MAP :  0.65\n"
     ]
    }
   ],
   "source": [
    "eval = pd.read_excel(\"data/evaluation_2016_2.xlsx\")\n",
    "emotions = ['Anger','Anticipation','Disgust','Fear','Joy','Sadness','Surprise','Trust','Love']\n",
    "epsilon = 0.00000001\n",
    "Precision(eval,emotions,epsilon)\n",
    "Recall(eval, emotions,epsilon )\n",
    "MAP(eval,'Anticipation')\n",
    "MAP(eval,'Joy')\n",
    "MAP(eval,'Sadness')\n",
    "MAP(eval,'Love')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\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": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}