NLP Project.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import nltk\n",
    "import sys\n",
    "import re\n",
    "import numpy as np\n",
    "from sklearn.feature_extraction import DictVectorizer\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.ensemble import GradientBoostingRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Read train data from files into 4 vocabularies - one per emotion\n",
    "\n",
    "anger = {}\n",
    "with open(\".\\EI-reg-En-train\\EI-reg-En-anger-train.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        anger[key] = float(val)\n",
    "        \n",
    "fear = {}\n",
    "with open(\".\\EI-reg-En-train\\EI-reg-En-fear-train.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        fear[key] = float(val)\n",
    "        \n",
    "joy = {}\n",
    "with open(\".\\EI-reg-En-train\\EI-reg-En-joy-train.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        joy[key] = float(val)\n",
    "        \n",
    "sadness = {}\n",
    "with open(\".\\EI-reg-En-train\\EI-reg-En-sadness-train.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        sadness[key] = float(val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x28e1c5f5518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x28e1b89f128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nr5V42QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x28e1bd340f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAD49JREFUeJzt3X+s3XV9x/HnS1AXJxu4XgwBuoIpZmi2qjfExehwOIdk\nAV2mo5mKjlg1suyHWeaPZBj3z6IyEjOHK7EBjVZwDG0WnBLiZFus81awKyoTsGKlaa/g0AXHBrz3\nx/l2HOtt7+n5fs+97YfnIzm53+/nfH+888ntq5/7Od/v96SqkCS160mrXYAkabYMeklqnEEvSY0z\n6CWpcQa9JDXOoJekxhn0ktQ4g16SGmfQS1Ljjl/tAgDWrFlT69atW+0yJOmYsmPHju9X1dxy2x0V\nQb9u3ToWFhZWuwxJOqYk+c4k2zl1I0mNM+glqXEGvSQ1zqCXpMYtG/RJtiTZn2TXWNt1SW7vXruT\n3N61r0vy47H3PjzL4iVJy5vkqptrgL8GPnqgoap+98BykiuAB8e2v7uqNgxVoCSpn2WDvqpuTbJu\nqfeSBHgN8OvDliVJGkrfOfoXA/uq6ltjbWckuS3JF5O8uOfxJUk99b1haiOwdWx9L7C2qu5P8gLg\n00meU1U/PHjHJJuATQBr167tWYYk6VCmHtEnOR74beC6A21V9XBV3d8t7wDuBs5aav+q2lxV81U1\nPze37B28OsYkw74kTa/P1M3LgG9W1Z4DDUnmkhzXLZ8JrAfu6VeiJKmPSS6v3Ap8CXh2kj1JLu3e\nupifnLYBeAmwM8nXgL8D3lJVDwxZsCTpyExy1c3GQ7S/YYm2G4Ab+pclSRqKd8ZKUuMMeklqnEEv\nSY0z6CWpcQa9JDXOoJekxhn0ktQ4g16SGmfQS1LjDHpJalzfxxSrET4hUmqXI3pJapxBL0mNM+gl\nqXEGvSQ1zqCXpMZ51Y2esIa+0qhq2ONJQ3FEL0mNc0SvY4LX+UvTc0QvSY0z6CWpccsGfZItSfYn\n2TXW9p4k30tye/e6YOy9dya5K8mdSX5zVoVLkiYzyYj+GuD8JdqvrKoN3esmgCRnAxcDz+n2+Zsk\nxw1VrCTpyC0b9FV1K/DAhMe7CPhkVT1cVd8G7gLO6VGfJKmnPnP0lyXZ2U3tnNS1nQp8d2ybPV3b\nT0myKclCkoXFxcUeZUiSDmfaoL8KeBawAdgLXNG1L3UR3JK3kVTV5qqar6r5ubm5Kct4YkqGf0lq\n11RBX1X7qurRqnoMuJrHp2f2AKePbXoacF+/EiVJfUwV9ElOGVt9FXDgipxtwMVJnprkDGA98G/9\nSpSODf6VpaPVsnfGJtkKnAusSbIHuBw4N8kGRtMyu4E3A1TVHUmuB74OPAK8raoenU3pkqRJpI6C\nJzHNz8/XwsLCapdxzHC098RwFPzT1FEuyY6qml9uO++MlaTGGfSS1DiDXpIaZ9BLUuMMeklqnEEv\nSY0z6CWpcX6VoHSU8svLNRRH9JLUOINekhpn0EtS4wx6SWqcQS9JjTPoJalxBr0kNc6gl6TGGfSS\n1DiDXpIaZ9BLUuMMeklq3LJBn2RLkv1Jdo21vT/JN5PsTHJjkhO79nVJfpzk9u714VkWL0la3iQj\n+muA8w9quxl4blX9MvAfwDvH3ru7qjZ0r7cMU6YkaVrLBn1V3Qo8cFDb56vqkW51O3DaDGqTJA1g\niDn63wc+O7Z+RpLbknwxyYsHOL4kqYdeXzyS5N3AI8DHu6a9wNqquj/JC4BPJ3lOVf1wiX03AZsA\n1q5d26cMSdJhTD2iT3IJ8FvA71WNvrumqh6uqvu75R3A3cBZS+1fVZurar6q5ufm5qYtQ5K0jKmC\nPsn5wJ8BF1bVQ2Ptc0mO65bPBNYD9wxRqCRpOstO3STZCpwLrEmyB7ic0VU2TwVuzuiLLbd3V9i8\nBHhvkkeAR4G3VNUDSx5YkrQilg36qtq4RPNHDrHtDcANfYuSJA3HO2MlqXEGvSQ1zqCXpMYZ9JLU\nOINekhrX685YSceO0ZXQwxrdKqmjnSN6SWqcQS9JjTPoJalxBr0kNc6gl6TGGfSS1DiDXpIaZ9BL\nUuMMeklqnEEvSY0z6CWpcQa9JDXOoJekxhn0ktQ4g16SGjdR0CfZkmR/kl1jbc9IcnOSb3U/T+ra\nk+SDSe5KsjPJ82dVvCRpeZOO6K8Bzj+o7R3ALVW1HrilWwd4BbC+e20CrupfpiRpWhMFfVXdCjxw\nUPNFwLXd8rXAK8faP1oj24ETk5wyRLGSpCPXZ47+mVW1F6D7eXLXfirw3bHt9nRtkqRVMIsPY5f6\nZsqf+mbJJJuSLCRZWFxcnEEZkiToF/T7DkzJdD/3d+17gNPHtjsNuO/gnatqc1XNV9X83NxcjzIk\nSYfTJ+i3AZd0y5cAnxlrf3139c0LgQcPTPFIklbe8ZNslGQrcC6wJske4HLgL4Hrk1wK3Au8utv8\nJuAC4C7gIeCNA9csSToCEwV9VW08xFvnLbFtAW/rU5QkaTjeGStJjTPoJalxBr0kNc6gl6TGGfSS\n1DiDXpIaZ9BLUuMMeklqnEEvSY0z6CWpcQa9JDXOoJekxhn0ktQ4g16SGmfQS1LjDHpJapxBL0mN\nM+glqXEGvSQ1zqCXpMYZ9JLUuOOn3THJs4HrxprOBP4cOBF4E7DYtb+rqm6aukJJUi9TB31V3Qls\nAEhyHPA94EbgjcCVVfWBQSqUJPUyddAf5Dzg7qr6TpKBDtkOu0TSahpqjv5iYOvY+mVJdibZkuSk\ngc4hSZpC76BP8hTgQuBTXdNVwLMYTevsBa44xH6bkiwkWVhcXFxqE0nSAIYY0b8C+GpV7QOoqn1V\n9WhVPQZcDZyz1E5Vtbmq5qtqfm5uboAyJElLGSLoNzI2bZPklLH3XgXsGuAckqQp9fowNsnTgN8A\n3jzW/L4kG4ACdh/0niRphfUK+qp6CPiFg9pe16siSdKgvDNWkhpn0EtS4wx6SWqcQS9JjTPoJalx\nBr0kNc6gl6TGGfSS1LihHlPcFB8rLKkljuglqXEGvSQ1zqCXpMY5Ry9pakN/nlU17PE04ohekhrn\niF7SUcO/EGbDEb0kNc6gl6TGGfSS1DiDXpIaZ9BLUuMMeklqXO/LK5PsBn4EPAo8UlXzSZ4BXAes\nA3YDr6mqH/Q9lyTpyA01on9pVW2oqvlu/R3ALVW1HrilW5ckrYJZTd1cBFzbLV8LvHJG55EkLWOI\noC/g80l2JNnUtT2zqvYCdD9PPninJJuSLCRZWFxcHKAMSdJShngEwouq6r4kJwM3J/nmJDtV1WZg\nM8D8/Lw3KkvSjPQe0VfVfd3P/cCNwDnAviSnAHQ/9/c9jyRpOr2CPsnPJjnhwDLwcmAXsA24pNvs\nEuAzfc4jSZpe36mbZwI3ZvTIueOBT1TVPyb5CnB9kkuBe4FX9zyPJGlKvYK+qu4BfmWJ9vuB8/oc\nW5I0DO+MlaTGGfSS1DiDXpIaZ9BLUuMMeklqnEEvSY0z6CWpcQa9JDXOoJekxhn0ktQ4g16SGmfQ\nS1LjDHpJapxBL0mNM+glqXFDfGfsqht974kkaSlNBL0kLWUWg8Cq4Y85a07dSFLjDHpJapxBL0mN\nmzrok5ye5AtJvpHkjiR/2LW/J8n3ktzevS4YrlxJ0pHq82HsI8Dbq+qrSU4AdiS5uXvvyqr6QP/y\nJEl9TR30VbUX2Nst/yjJN4BThypMkjSMQebok6wDngd8uWu6LMnOJFuSnDTEOSRJ0+kd9EmeDtwA\n/FFV/RC4CngWsIHRiP+KQ+y3KclCkoXFxcW+ZUiSDqFX0Cd5MqOQ/3hV/T1AVe2rqker6jHgauCc\npfatqs1VNV9V83Nzc33KkCQdRp+rbgJ8BPhGVf3VWPspY5u9Ctg1fXmSpL76XHXzIuB1wL8nub1r\nexewMckGoIDdwJt7VShJ6qXPVTf/Aiz1JImbpi9HkjQ074yVpMYZ9JLUOINekhpn0EtS4wx6SWqc\nQS9JjTPoJalxBr0kNc6gl6TGGfSS1DiDXpIaZ9BLUuMMeklqnEEvSY0z6CWpcX2+eESSnnCy1Ldw\n9FA17PGW4ohekhpn0EtS4wx6SWqcQS9JjTPoJalxMwv6JOcnuTPJXUneMavzSJIObyZBn+Q44EPA\nK4CzgY1Jzp7FuSRJhzerEf05wF1VdU9V/Q/wSeCiGZ1LknQYswr6U4Hvjq3v6dokSStsVnfGLnXv\n2E/c/5VkE7CpW/2vJHfOqJbVtgb4/moXcRSwHx5nX4zYD0DSqx9+cZKNZhX0e4DTx9ZPA+4b36Cq\nNgObZ3T+o0aShaqaX+06Vpv98Dj7YsR+GFmJfpjV1M1XgPVJzkjyFOBiYNuMziVJOoyZjOir6pEk\nlwGfA44DtlTVHbM4lyTp8Gb29Mqqugm4aVbHP4Y0Pz01IfvhcfbFiP0wMvN+SK3EMzIlSavGRyBI\nUuMM+oEs98iHJH+S5OtJdia5JclEl0UdayZ99EWS30lSSZq86mKSfkjymu534o4kn1jpGlfCBP8u\n1ib5QpLbun8bF6xGnbOWZEuS/Ul2HeL9JPlg1087kzx/0AKqylfPF6MPnO8GzgSeAnwNOPugbV4K\nPK1bfitw3WrXvRr90G13AnArsB2YX+26V+n3YT1wG3BSt37yate9Sv2wGXhrt3w2sHu1655RX7wE\neD6w6xDvXwB8ltE9SC8Evjzk+R3RD2PZRz5U1Req6qFudTujewtaM+mjL/4CeB/w3ytZ3AqapB/e\nBHyoqn4AUFX7V7jGlTBJPxTwc93yz3PQ/TatqKpbgQcOs8lFwEdrZDtwYpJThjq/QT+MI33kw6WM\n/vduzbL9kOR5wOlV9Q8rWdgKm+T34SzgrCT/mmR7kvNXrLqVM0k/vAd4bZI9jK7S+4OVKe2oM9PH\nxvjl4MNY9pEP/79h8lpgHvi1mVa0Og7bD0meBFwJvGGlClolk/w+HM9o+uZcRn/d/XOS51bVf864\ntpU0ST9sBK6pqiuS/Crwsa4fHpt9eUeViTNkGo7oh7HsIx8AkrwMeDdwYVU9vEK1raTl+uEE4LnA\nPyXZzWgucluDH8hO8vuwB/hMVf1vVX0buJNR8Ldkkn64FLgeoKq+BPwMo2fgPNFMlCHTMuiHsewj\nH7opi79lFPItzsfCMv1QVQ9W1ZqqWldV6xh9VnFhVS2sTrkzM8kjQD7N6AN6kqxhNJVzz4pWOXuT\n9MO9wHkASX6JUdAvrmiVR4dtwOu7q29eCDxYVXuHOrhTNwOoQzzyIcl7gYWq2ga8H3g68KkkAPdW\n1YWrVvQMTNgPzZuwHz4HvDzJ14FHgT+tqvtXr+rhTdgPbweuTvLHjKYq3lDdZSgtSbKV0TTdmu7z\niMuBJwNU1YcZfT5xAXAX8BDwxkHP32CfSpLGOHUjSY0z6CWpcQa9JDXOoJekxhn0ktQ4g16SGmfQ\nS1LjDHpJatz/ARcTDJ4ive1VAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x28e1c22cb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "plt.hist([anger[k] for k in anger.keys()], bins=15, color='red')\n",
    "plt.show()\n",
    "\n",
    "plt.hist([fear[k] for k in fear.keys()], bins=15, color='violet')\n",
    "plt.show()\n",
    "\n",
    "plt.hist([joy[k] for k in joy.keys()], bins=15, color='yellow')\n",
    "plt.show()\n",
    "\n",
    "plt.hist([sadness[k] for k in sadness.keys()], bins=15, color='blue')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Read dev data from files into 4 vocabularies - one per emotion\n",
    "\n",
    "anger_dev = {}\n",
    "with open(\".\\EI-reg-En-dev\\EI-reg-En-anger-dev.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        anger_dev[key] = float(val)\n",
    "        \n",
    "fear_dev = {}\n",
    "with open(\".\\EI-reg-En-dev\\EI-reg-En-fear-dev.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        fear_dev[key] = float(val)\n",
    "        \n",
    "joy_dev = {}\n",
    "with open(\".\\EI-reg-En-dev\\EI-reg-En-joy-dev.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        joy_dev[key] = float(val)\n",
    "        \n",
    "sadness_dev = {}\n",
    "with open(\".\\EI-reg-En-dev\\EI-reg-En-sadness-dev.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        (idn, key, emotion, val) = line.split(\"\\t\", 4)\n",
    "        val = val[:len(val)-2]\n",
    "        sadness_dev[key] = float(val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Data tokenization\n",
    "\n",
    "def tokenize(s):\n",
    "        return tokens_re.findall(s)\n",
    "\n",
    "def preprocess(s, lowercase=False):\n",
    "    tokens = tokenize(s)\n",
    "    if lowercase:\n",
    "        for token in tokens:\n",
    "            if emoticon_re.search(token):\n",
    "                tokens.append(token)\n",
    "            else:\n",
    "                tokens.append(token.lower())\n",
    "    return tokens\n",
    "\n",
    "emoticons_str = r\"\"\"\n",
    "    (?:\n",
    "        [:=;] \n",
    "        [oO\\-]? \n",
    "        [D\\)\\]\\(\\]/\\\\OpP] \n",
    "    )\"\"\"\n",
    "regex_str = [\n",
    "    emoticons_str, # old-style emojis\n",
    "    r'<[^>]+>', # HTML-tags\n",
    "    r'(?:@[\\w_]+)', # @-mentions of other users\n",
    "    r\"(?:\\#+[\\w_]+[\\w\\'_\\-]*[\\w_]+)\", # hashtags\n",
    "    r'http[s]?://(?:[a-z]|[0-9]|[$-_@.&amp;+]|[!*\\(\\),]|(?:%[0-9a-f][0-9a-f]))+', # URLs\n",
    "    r'(?:(?:\\d+,?)+(?:\\.?\\d+)?)', # numbers\n",
    "    r\"(?:[a-z][a-z'\\-_]+[a-z])\", # words with - and ' symbols\n",
    "    r'(?:[\\w_]+)', # other words\n",
    "    r'(?:\\S)' # other symbols\n",
    "]\n",
    "\n",
    "#if data with emojis will give bad result, we could skip emojis using this pattern\n",
    "#emoji_pattern = re.compile(\"[\"\n",
    "#        u\"\\U0001F600-\\U0001F64F\"  # emoticons\n",
    "#        u\"\\U0001F300-\\U0001F5FF\"  # symbols & pictographs\n",
    "#        u\"\\U0001F680-\\U0001F6FF\"  # transport & map symbols\n",
    "#        u\"\\U0001F1E0-\\U0001F1FF\"  # flags (iOS)\n",
    "#                           \"]+\", flags=re.UNICODE)\n",
    "\n",
    "marks = ['.', ',', '-', '/', '…', '’', '—', '\\\\', '\\'', '\\\"', '!', '!!', '@', '%', '^', '?', '&', '*', '(', ')', '_', 'n', '#', '+', '=', ';', ':', '$', '£', '\\`', '~', '<', '>', ':)', ';)']\n",
    "\n",
    "from nltk.corpus import stopwords\n",
    "stop_words = set(stopwords.words('english'))\n",
    "\n",
    "tokens_re = re.compile(r'('+'|'.join(regex_str)+')', re.VERBOSE | re.IGNORECASE)\n",
    "emoticon_re = re.compile(r'^'+emoticons_str+'$', re.VERBOSE | re.IGNORECASE)\n",
    "\n",
    "def filter_dict(d, tokens, indexes):\n",
    "    for k in d.keys():\n",
    "        k = preprocess(k)\n",
    "        new_key = []\n",
    "        for i in k:\n",
    "            if i not in stop_words and i not in marks:\n",
    "                i = re.sub(r'(?:@[\\w_]+)', \"\", i) # delete @-mentions\n",
    "                i = re.sub(r'<[^>]+>', \"\", i) # delete HTML tags\n",
    "                i = re.sub(r'http[s]?://(?:[a-z]|[0-9]|[$-_@.&amp;+]|[!*\\(\\),]|(?:%[0-9a-f][0-9a-f]))+', \"\", i) # delete URLs\n",
    "                i = re.sub(r'(?:(?:\\d+,?)+(?:\\.?\\d+)?)', \"\", i) # delete numbers\n",
    "                #i = emoji_pattern.sub(r'', i) # delete new emojis\n",
    "                #i = re.sub(emoticons_str, '', i) # delete old emoticons\n",
    "                if len(i) > 1 and i[0] == '#':\n",
    "                    i = i[1:]\n",
    "                if i is not '':\n",
    "                    new_key.append(i)            \n",
    "        tokens.append(new_key)        \n",
    "    for v in d.values():\n",
    "        indexes.append(v)\n",
    "\n",
    "#For train data        \n",
    "anger_tokens = []\n",
    "anger_ind = []\n",
    "filter_dict(anger, anger_tokens, anger_ind)\n",
    "fear_tokens = []\n",
    "fear_ind = []\n",
    "filter_dict(fear, fear_tokens, fear_ind)\n",
    "joy_tokens = []\n",
    "joy_ind = []\n",
    "filter_dict(joy, joy_tokens, joy_ind)\n",
    "sadness_tokens = []\n",
    "sadness_ind = []\n",
    "filter_dict(sadness, sadness_tokens, sadness_ind)\n",
    "\n",
    "#For dev data\n",
    "anger_tokens_dev = []\n",
    "anger_ind_dev = []\n",
    "filter_dict(anger_dev, anger_tokens_dev, anger_ind_dev)\n",
    "fear_tokens_dev = []\n",
    "fear_ind_dev = []\n",
    "filter_dict(fear_dev, fear_tokens_dev, fear_ind_dev)\n",
    "joy_tokens_dev = []\n",
    "joy_ind_dev = []\n",
    "filter_dict(joy_dev, joy_tokens_dev, joy_ind_dev)\n",
    "sadness_tokens_dev = []\n",
    "sadness_ind_dev = []\n",
    "filter_dict(sadness_dev, sadness_tokens_dev, sadness_ind_dev)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gensim.models\n",
    "Vec = gensim.models.KeyedVectors.load_word2vec_format('.\\GoogleNews-vectors-negative300.bin\\\\GoogleNews-vectors-negative300.bin', binary=True)\n",
    "model = Vec.wv\n",
    "del Vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_feat(data):\n",
    "    feat = []\n",
    "    for i in range(len(data)):\n",
    "        f = []\n",
    "        dat = data[i] # (data[i]).split(\" \")\n",
    "        temp = []\n",
    "        for w in dat:\n",
    "            if w in model:\n",
    "                temp.append(model[w])\n",
    "        if len(temp) == 0:\n",
    "            temp.append(model['cat'])\n",
    "        f = list(((np.array(temp)).mean(0)))\n",
    "        f += list((np.median((np.array(temp)), 0)))\n",
    "        f += list((np.array(temp)).max(0))\n",
    "        f += list((np.array(temp)).min(0))\n",
    "        f += list((np.percentile(temp, 25, 0)))\n",
    "        f += list((np.percentile(temp, 75, 0)))\n",
    "        n = [len(k) for k in dat]\n",
    "        f.append(len(dat))\n",
    "        f.append(sum(n))\n",
    "        f.append(sum(n) / len(dat))\n",
    "        feat.append(f)\n",
    "    feat = np.array(feat)\n",
    "    feat = np.reshape(feat, (len(data), 1803))\n",
    "    return feat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Feature extracting process\n",
    "\n",
    "#For train data\n",
    "anger_dicts, anger_indexes = get_feat(anger_tokens), anger_ind\n",
    "fear_dicts, fear_indexes = get_feat(fear_tokens), fear_ind\n",
    "joy_dicts, joy_indexes = get_feat(joy_tokens), joy_ind\n",
    "sadness_dicts, sadness_indexes = get_feat(sadness_tokens), sadness_ind\n",
    "\n",
    "#For dev data\n",
    "anger_dicts_dev, anger_indexes_dev = get_feat(anger_tokens_dev), anger_ind_dev\n",
    "fear_dicts_dev, fear_indexes_dev = get_feat(fear_tokens_dev), fear_ind_dev\n",
    "joy_dicts_dev, joy_indexes_dev = get_feat(joy_tokens_dev), joy_ind_dev\n",
    "sadness_dicts_dev, sadness_indexes_dev = get_feat(sadness_tokens_dev), sadness_ind_dev\n",
    "\n",
    "train_anger = anger_dicts\n",
    "dev_anger = anger_dicts_dev\n",
    "train_fear = fear_dicts\n",
    "dev_fear = fear_dicts_dev\n",
    "train_joy = joy_dicts\n",
    "dev_joy = joy_dicts_dev\n",
    "train_sadness = sadness_dicts\n",
    "dev_sadness = sadness_dicts_dev"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVR with RBF kernel: dev set RMSE for anger is 0.0100 while c is 0.01\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0090 while c is 0.01\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0110 while c is 0.01\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0093 while c is 0.01\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0100 while c is 0.02\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0090 while c is 0.02\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0110 while c is 0.02\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0093 while c is 0.02\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0099 while c is 0.05\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0089 while c is 0.05\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0110 while c is 0.05\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0093 while c is 0.05\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0098 while c is 0.10\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 0.10\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0109 while c is 0.10\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 0.10\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 0.20\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 0.20\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0108 while c is 0.20\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 0.20\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 0.50\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 0.50\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 0.50\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 0.50\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 1.00\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 1.00\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 1.00\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 1.00\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 2.00\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 2.00\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 2.00\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 2.00\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 5.00\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 5.00\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 5.00\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 5.00\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 10.00\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 10.00\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 10.00\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 10.00\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 20.00\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 20.00\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 20.00\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 20.00\n",
      "SVR with RBF kernel: dev set RMSE for anger is 0.0097 while c is 50.00\n",
      "SVR with RBF kernel: dev set RMSE for fear is 0.0088 while c is 50.00\n",
      "SVR with RBF kernel: dev set RMSE for joy is 0.0107 while c is 50.00\n",
      "SVR with RBF kernel: dev set RMSE for sadness is 0.0092 while c is 50.00\n"
     ]
    }
   ],
   "source": [
    "#SVM model with rbf kernel\n",
    "\n",
    "def model_svm_rbf(trainX, trainY, devX, devY, c):\n",
    "    model = SVR(kernel='rbf', C=c, gamma=0.1)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "c_values = [0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, 50]\n",
    "accuracy_anger_svm = []\n",
    "accuracy_fear_svm = []\n",
    "accuracy_joy_svm = []\n",
    "accuracy_sadness_svm = []\n",
    "for c in c_values:\n",
    "    acc_anger = model_svm_rbf(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_svm.append(acc_anger)\n",
    "    print(\"SVR with RBF kernel: dev set RMSE for anger is %.4f\" % acc_anger, \"while c is %.2f\" % c)\n",
    "    acc_fear = model_svm_rbf(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_svm.append(acc_fear)\n",
    "    print(\"SVR with RBF kernel: dev set RMSE for fear is %.4f\" % acc_fear, \"while c is %.2f\" % c)\n",
    "    acc_joy = model_svm_rbf(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_svm.append(acc_joy)\n",
    "    print(\"SVR with RBF kernel: dev set RMSE for joy is %.4f\" % acc_joy, \"while c is %.2f\" % c)\n",
    "    acc_sadness = model_svm_rbf(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_svm.append(acc_sadness)\n",
    "    print(\"SVR with RBF kernel: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while c is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVR with linear kernel: dev set RMSE for anger is 0.0088 while c is 0.01\n",
      "SVR with linear kernel: dev set RMSE for fear is 0.0079 while c is 0.01\n",
      "SVR with linear kernel: dev set RMSE for joy is 0.0103 while c is 0.01\n",
      "SVR with linear kernel: dev set RMSE for sadness is 0.0084 while c is 0.01\n",
      "SVR with linear kernel: dev set RMSE for anger is 0.0100 while c is 0.10\n",
      "SVR with linear kernel: dev set RMSE for fear is 0.0092 while c is 0.10\n",
      "SVR with linear kernel: dev set RMSE for joy is 0.0118 while c is 0.10\n",
      "SVR with linear kernel: dev set RMSE for sadness is 0.0096 while c is 0.10\n",
      "SVR with linear kernel: dev set RMSE for anger is 0.0116 while c is 1.00\n",
      "SVR with linear kernel: dev set RMSE for fear is 0.0125 while c is 1.00\n",
      "SVR with linear kernel: dev set RMSE for joy is 0.0134 while c is 1.00\n",
      "SVR with linear kernel: dev set RMSE for sadness is 0.0110 while c is 1.00\n",
      "SVR with linear kernel: dev set RMSE for anger is 0.0156 while c is 10.00\n",
      "SVR with linear kernel: dev set RMSE for fear is 0.0167 while c is 10.00\n",
      "SVR with linear kernel: dev set RMSE for joy is 0.0148 while c is 10.00\n",
      "SVR with linear kernel: dev set RMSE for sadness is 0.0133 while c is 10.00\n"
     ]
    }
   ],
   "source": [
    "#SVM model with linear kernel\n",
    "\n",
    "def model_svm_lin(trainX, trainY, devX, devY, c):\n",
    "    model = SVR(kernel = 'linear', C = c)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "c_values = [0.01, 0.1, 1, 10]\n",
    "accuracy_anger_svm = []\n",
    "accuracy_fear_svm = []\n",
    "accuracy_joy_svm = []\n",
    "accuracy_sadness_svm = []\n",
    "for c in c_values:\n",
    "    acc_anger = model_svm_lin(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_svm.append(acc_anger)\n",
    "    print(\"SVR with linear kernel: dev set RMSE for anger is %.4f\" % acc_anger, \"while c is %.2f\" % c)\n",
    "    acc_fear = model_svm_lin(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_svm.append(acc_fear)\n",
    "    print(\"SVR with linear kernel: dev set RMSE for fear is %.4f\" % acc_fear, \"while c is %.2f\" % c)\n",
    "    acc_joy = model_svm_lin(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_svm.append(acc_joy)\n",
    "    print(\"SVR with linear kernel: dev set RMSE for joy is %.4f\" % acc_joy, \"while c is %.2f\" % c)\n",
    "    acc_sadness = model_svm_lin(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_svm.append(acc_sadness)\n",
    "    print(\"SVR with linear kernel: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while c is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVR with polynomial kernel: dev set RMSE for anger is 0.0100 while c is 0.01\n",
      "SVR with polynomial kernel: dev set RMSE for fear is 0.0089 while c is 0.01\n",
      "SVR with polynomial kernel: dev set RMSE for joy is 0.0109 while c is 0.01\n",
      "SVR with polynomial kernel: dev set RMSE for sadness is 0.0092 while c is 0.01\n",
      "SVR with polynomial kernel: dev set RMSE for anger is 0.0097 while c is 0.10\n",
      "SVR with polynomial kernel: dev set RMSE for fear is 0.0085 while c is 0.10\n",
      "SVR with polynomial kernel: dev set RMSE for joy is 0.0104 while c is 0.10\n",
      "SVR with polynomial kernel: dev set RMSE for sadness is 0.0090 while c is 0.10\n",
      "SVR with polynomial kernel: dev set RMSE for anger is 0.0089 while c is 1.00\n",
      "SVR with polynomial kernel: dev set RMSE for fear is 0.0079 while c is 1.00\n",
      "SVR with polynomial kernel: dev set RMSE for joy is 0.0097 while c is 1.00\n",
      "SVR with polynomial kernel: dev set RMSE for sadness is 0.0084 while c is 1.00\n",
      "SVR with polynomial kernel: dev set RMSE for anger is 0.0089 while c is 10.00\n",
      "SVR with polynomial kernel: dev set RMSE for fear is 0.0082 while c is 10.00\n",
      "SVR with polynomial kernel: dev set RMSE for joy is 0.0107 while c is 10.00\n",
      "SVR with polynomial kernel: dev set RMSE for sadness is 0.0086 while c is 10.00\n"
     ]
    }
   ],
   "source": [
    "#SVM model with poly kernel\n",
    "\n",
    "def model_svm_poly(trainX, trainY, devX, devY, c):\n",
    "    model = SVR(kernel='poly', C = c, degree = 2)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "c_values = [0.01, 0.1, 1, 10]\n",
    "accuracy_anger_svm = []\n",
    "accuracy_fear_svm = []\n",
    "accuracy_joy_svm = []\n",
    "accuracy_sadness_svm = []\n",
    "for c in c_values:\n",
    "    acc_anger = model_svm_poly(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_svm.append(acc_anger)\n",
    "    print(\"SVR with polynomial kernel: dev set RMSE for anger is %.4f\" % acc_anger, \"while c is %.2f\" % c)\n",
    "    acc_fear = model_svm_poly(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_svm.append(acc_fear)\n",
    "    print(\"SVR with polynomial kernel: dev set RMSE for fear is %.4f\" % acc_fear, \"while c is %.2f\" % c)\n",
    "    acc_joy = model_svm_poly(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_svm.append(acc_joy)\n",
    "    print(\"SVR with polynomial kernel: dev set RMSE for joy is %.4f\" % acc_joy, \"while c is %.2f\" % c)\n",
    "    acc_sadness = model_svm_poly(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_svm.append(acc_sadness)\n",
    "    print(\"SVR with polynomial kernel: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while c is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision tree: dev set RMSE for anger is 0.0125 while feature number is 10.00\n",
      "Decision tree: dev set RMSE for fear is 0.0118 while feature number is 10.00\n",
      "Decision tree: dev set RMSE for joy is 0.0155 while feature number is 10.00\n",
      "Decision tree: dev set RMSE for sadness is 0.0131 while feature number is 10.00\n",
      "Decision tree: dev set RMSE for anger is 0.0116 while feature number is 110.00\n",
      "Decision tree: dev set RMSE for fear is 0.0126 while feature number is 110.00\n",
      "Decision tree: dev set RMSE for joy is 0.0150 while feature number is 110.00\n",
      "Decision tree: dev set RMSE for sadness is 0.0128 while feature number is 110.00\n",
      "Decision tree: dev set RMSE for anger is 0.0117 while feature number is 210.00\n",
      "Decision tree: dev set RMSE for fear is 0.0133 while feature number is 210.00\n",
      "Decision tree: dev set RMSE for joy is 0.0161 while feature number is 210.00\n",
      "Decision tree: dev set RMSE for sadness is 0.0124 while feature number is 210.00\n",
      "Decision tree: dev set RMSE for anger is 0.0124 while feature number is 310.00\n",
      "Decision tree: dev set RMSE for fear is 0.0126 while feature number is 310.00\n",
      "Decision tree: dev set RMSE for joy is 0.0154 while feature number is 310.00\n",
      "Decision tree: dev set RMSE for sadness is 0.0127 while feature number is 310.00\n",
      "Decision tree: dev set RMSE for anger is 0.0120 while feature number is 410.00\n",
      "Decision tree: dev set RMSE for fear is 0.0123 while feature number is 410.00\n",
      "Decision tree: dev set RMSE for joy is 0.0151 while feature number is 410.00\n",
      "Decision tree: dev set RMSE for sadness is 0.0128 while feature number is 410.00\n",
      "Decision tree: dev set RMSE for anger is 0.0122 while feature number is 510.00\n",
      "Decision tree: dev set RMSE for fear is 0.0119 while feature number is 510.00\n",
      "Decision tree: dev set RMSE for joy is 0.0157 while feature number is 510.00\n",
      "Decision tree: dev set RMSE for sadness is 0.0128 while feature number is 510.00\n"
     ]
    }
   ],
   "source": [
    "#Decision tree model\n",
    "\n",
    "def model_dt(trainX, trainY, devX, devY, c):\n",
    "    model = DecisionTreeRegressor(max_features = c)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "features_number = np.arange(10,521,100)\n",
    "accuracy_anger_dt = []\n",
    "accuracy_fear_dt = []\n",
    "accuracy_joy_dt = []\n",
    "accuracy_sadness_dt = []\n",
    "for c in features_number:\n",
    "    acc_anger = model_dt(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_dt.append(acc_anger)\n",
    "    print(\"Decision tree: dev set RMSE for anger is %.4f\" % acc_anger, \"while feature number is %.2f\" % c)\n",
    "    acc_fear = model_dt(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_dt.append(acc_fear)\n",
    "    print(\"Decision tree: dev set RMSE for fear is %.4f\" % acc_fear, \"while feature number is %.2f\" % c)\n",
    "    acc_joy = model_dt(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_dt.append(acc_joy)\n",
    "    print(\"Decision tree: dev set RMSE for joy is %.4f\" % acc_joy, \"while feature number is %.2f\" % c)\n",
    "    acc_sadness = model_dt(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_dt.append(acc_sadness)\n",
    "    print(\"Decision tree: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while feature number is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "k-NearestNeighbors: dev set RMSE for anger is 0.0110 while number of neighbors is 2.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0107 while number of neighbors is 2.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0127 while number of neighbors is 2.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0118 while number of neighbors is 2.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0102 while number of neighbors is 4.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0099 while number of neighbors is 4.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0114 while number of neighbors is 4.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0103 while number of neighbors is 4.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0101 while number of neighbors is 6.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0094 while number of neighbors is 6.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0112 while number of neighbors is 6.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0099 while number of neighbors is 6.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0099 while number of neighbors is 8.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0093 while number of neighbors is 8.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0111 while number of neighbors is 8.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0097 while number of neighbors is 8.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0098 while number of neighbors is 10.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0090 while number of neighbors is 10.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0108 while number of neighbors is 10.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0096 while number of neighbors is 10.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0098 while number of neighbors is 12.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0090 while number of neighbors is 12.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0109 while number of neighbors is 12.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0096 while number of neighbors is 12.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0098 while number of neighbors is 14.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0090 while number of neighbors is 14.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0110 while number of neighbors is 14.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0096 while number of neighbors is 14.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0098 while number of neighbors is 16.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0089 while number of neighbors is 16.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0110 while number of neighbors is 16.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0095 while number of neighbors is 16.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0097 while number of neighbors is 18.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0089 while number of neighbors is 18.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0110 while number of neighbors is 18.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0095 while number of neighbors is 18.00\n",
      "k-NearestNeighbors: dev set RMSE for anger is 0.0098 while number of neighbors is 20.00\n",
      "k-NearestNeighbors: dev set RMSE for fear is 0.0089 while number of neighbors is 20.00\n",
      "k-NearestNeighbors: dev set RMSE for joy is 0.0111 while number of neighbors is 20.00\n",
      "k-NearestNeighbors: dev set RMSE for sadness is 0.0095 while number of neighbors is 20.00\n"
     ]
    }
   ],
   "source": [
    "#k-NearestNeighbors Regressor\n",
    "\n",
    "def model_knn(trainX, trainY, devX, devY, c):\n",
    "    model = KNeighborsRegressor(n_neighbors = c)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "neighbors_number = np.arange(2,21,2)\n",
    "accuracy_anger_knn = []\n",
    "accuracy_fear_knn = []\n",
    "accuracy_joy_knn = []\n",
    "accuracy_sadness_knn = []\n",
    "for c in neighbors_number:\n",
    "    acc_anger = model_knn(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_knn.append(acc_anger)\n",
    "    print(\"k-NearestNeighbors: dev set RMSE for anger is %.4f\" % acc_anger, \"while number of neighbors is %.2f\" % c)\n",
    "    acc_fear = model_knn(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_knn.append(acc_fear)\n",
    "    print(\"k-NearestNeighbors: dev set RMSE for fear is %.4f\" % acc_fear, \"while number of neighbors is %.2f\" % c)\n",
    "    acc_joy = model_knn(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_knn.append(acc_joy)\n",
    "    print(\"k-NearestNeighbors: dev set RMSE for joy is %.4f\" % acc_joy, \"while number of neighbors is %.2f\" % c)\n",
    "    acc_sadness = model_knn(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_knn.append(acc_sadness)\n",
    "    print(\"k-NearestNeighbors: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while number of neighbors is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest: dev set RMSE for anger is 0.0098 while number of trees is 10.00\n",
      "Random Forest: dev set RMSE for fear is 0.0086 while feature number of trees is 10.00\n",
      "Random Forest: dev set RMSE for joy is 0.0104 while feature number of trees is 10.00\n",
      "Random Forest: dev set RMSE for sadness is 0.0092 while feature number of trees is 10.00\n",
      "Random Forest: dev set RMSE for anger is 0.0092 while number of trees is 60.00\n",
      "Random Forest: dev set RMSE for fear is 0.0083 while feature number of trees is 60.00\n",
      "Random Forest: dev set RMSE for joy is 0.0098 while feature number of trees is 60.00\n",
      "Random Forest: dev set RMSE for sadness is 0.0087 while feature number of trees is 60.00\n",
      "Random Forest: dev set RMSE for anger is 0.0093 while number of trees is 110.00\n",
      "Random Forest: dev set RMSE for fear is 0.0082 while feature number of trees is 110.00\n",
      "Random Forest: dev set RMSE for joy is 0.0098 while feature number of trees is 110.00\n",
      "Random Forest: dev set RMSE for sadness is 0.0086 while feature number of trees is 110.00\n",
      "Random Forest: dev set RMSE for anger is 0.0093 while number of trees is 160.00\n",
      "Random Forest: dev set RMSE for fear is 0.0082 while feature number of trees is 160.00\n",
      "Random Forest: dev set RMSE for joy is 0.0097 while feature number of trees is 160.00\n",
      "Random Forest: dev set RMSE for sadness is 0.0087 while feature number of trees is 160.00\n",
      "Random Forest: dev set RMSE for anger is 0.0093 while number of trees is 210.00\n",
      "Random Forest: dev set RMSE for fear is 0.0082 while feature number of trees is 210.00\n",
      "Random Forest: dev set RMSE for joy is 0.0098 while feature number of trees is 210.00\n",
      "Random Forest: dev set RMSE for sadness is 0.0086 while feature number of trees is 210.00\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-116-96cf192a96f5>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     14\u001b[0m \u001b[0maccuracy_sadness_rf\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     15\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mc\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mtrees_number\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 16\u001b[1;33m     \u001b[0macc_anger\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel_rf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtrain_anger\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0manger_indexes\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdev_anger\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0manger_indexes_dev\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     17\u001b[0m     \u001b[0maccuracy_anger_rf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0macc_anger\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     18\u001b[0m     \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"Random Forest: dev set RMSE for anger is %.4f\"\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0macc_anger\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"while number of trees is %.2f\"\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0mc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-116-96cf192a96f5>\u001b[0m in \u001b[0;36mmodel_rf\u001b[1;34m(trainX, trainY, devX, devY, c)\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;32mdef\u001b[0m \u001b[0mmodel_rf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtrainX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrainY\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdevX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdevY\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m     \u001b[0mmodel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mRandomForestRegressor\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mn_estimators\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 5\u001b[1;33m     \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtrainX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrainY\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      6\u001b[0m     \u001b[0mtrain_predict\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdevX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      7\u001b[0m     \u001b[0mtrain_acc\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlinalg\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnorm\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtrain_predict\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mdevY\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m/\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtrain_predict\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\ensemble\\forest.py\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, X, y, sample_weight)\u001b[0m\n\u001b[0;32m    325\u001b[0m                     \u001b[0mt\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtrees\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    326\u001b[0m                     verbose=self.verbose, class_weight=self.class_weight)\n\u001b[1;32m--> 327\u001b[1;33m                 for i, t in enumerate(trees))\n\u001b[0m\u001b[0;32m    328\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    329\u001b[0m             \u001b[1;31m# Collect newly grown trees\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\parallel.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, iterable)\u001b[0m\n\u001b[0;32m    777\u001b[0m             \u001b[1;31m# was dispatched. In particular this covers the edge\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    778\u001b[0m             \u001b[1;31m# case of Parallel used with an exhausted iterator.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 779\u001b[1;33m             \u001b[1;32mwhile\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdispatch_one_batch\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    780\u001b[0m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_iterating\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    781\u001b[0m             \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\parallel.py\u001b[0m in \u001b[0;36mdispatch_one_batch\u001b[1;34m(self, iterator)\u001b[0m\n\u001b[0;32m    623\u001b[0m                 \u001b[1;32mreturn\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    624\u001b[0m             \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 625\u001b[1;33m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_dispatch\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtasks\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    626\u001b[0m                 \u001b[1;32mreturn\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    627\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\parallel.py\u001b[0m in \u001b[0;36m_dispatch\u001b[1;34m(self, batch)\u001b[0m\n\u001b[0;32m    586\u001b[0m         \u001b[0mdispatch_timestamp\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    587\u001b[0m         \u001b[0mcb\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mBatchCompletionCallBack\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdispatch_timestamp\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 588\u001b[1;33m         \u001b[0mjob\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_backend\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mapply_async\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcallback\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mcb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    589\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_jobs\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mjob\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    590\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\_parallel_backends.py\u001b[0m in \u001b[0;36mapply_async\u001b[1;34m(self, func, callback)\u001b[0m\n\u001b[0;32m    109\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mapply_async\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfunc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcallback\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    110\u001b[0m         \u001b[1;34m\"\"\"Schedule a func to be run\"\"\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 111\u001b[1;33m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mImmediateResult\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    112\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mcallback\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    113\u001b[0m             \u001b[0mcallback\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\_parallel_backends.py\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, batch)\u001b[0m\n\u001b[0;32m    330\u001b[0m         \u001b[1;31m# Don't delay the application, to avoid keeping the input\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    331\u001b[0m         \u001b[1;31m# arguments in memory\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 332\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mresults\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    333\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    334\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\parallel.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    129\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    130\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 131\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mfunc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    132\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    133\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__len__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\externals\\joblib\\parallel.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    129\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    130\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 131\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mfunc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    132\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    133\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__len__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\ensemble\\forest.py\u001b[0m in \u001b[0;36m_parallel_build_trees\u001b[1;34m(tree, forest, X, y, sample_weight, tree_idx, n_trees, verbose, class_weight)\u001b[0m\n\u001b[0;32m    118\u001b[0m             \u001b[0mcurr_sample_weight\u001b[0m \u001b[1;33m*=\u001b[0m \u001b[0mcompute_sample_weight\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'balanced'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mindices\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    119\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 120\u001b[1;33m         \u001b[0mtree\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mcurr_sample_weight\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcheck_input\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    121\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    122\u001b[0m         \u001b[0mtree\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcheck_input\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\tree\\tree.py\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, X, y, sample_weight, check_input, X_idx_sorted)\u001b[0m\n\u001b[0;32m   1122\u001b[0m             \u001b[0msample_weight\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1123\u001b[0m             \u001b[0mcheck_input\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mcheck_input\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1124\u001b[1;33m             X_idx_sorted=X_idx_sorted)\n\u001b[0m\u001b[0;32m   1125\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1126\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\sklearn\\tree\\tree.py\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, X, y, sample_weight, check_input, X_idx_sorted)\u001b[0m\n\u001b[0;32m    360\u001b[0m                                            min_impurity_split)\n\u001b[0;32m    361\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 362\u001b[1;33m         \u001b[0mbuilder\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbuild\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtree_\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mX_idx_sorted\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    363\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    364\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mn_outputs_\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "#Random Forest Regressor\n",
    "\n",
    "def model_rf(trainX, trainY, devX, devY, c):\n",
    "    model = RandomForestRegressor(n_estimators = c)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "trees_number = np.arange(10, 211, 50)\n",
    "accuracy_anger_rf = []\n",
    "accuracy_fear_rf = []\n",
    "accuracy_joy_rf = []\n",
    "accuracy_sadness_rf = []\n",
    "for c in trees_number:\n",
    "    acc_anger = model_rf(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_rf.append(acc_anger)\n",
    "    print(\"Random Forest: dev set RMSE for anger is %.4f\" % acc_anger, \"while number of trees is %.2f\" % c)\n",
    "    acc_fear = model_rf(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_rf.append(acc_fear)\n",
    "    print(\"Random Forest: dev set RMSE for fear is %.4f\" % acc_fear, \"while feature number of trees is %.2f\" % c)\n",
    "    acc_joy = model_rf(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_rf.append(acc_joy)\n",
    "    print(\"Random Forest: dev set RMSE for joy is %.4f\" % acc_joy, \"while feature number of trees is %.2f\" % c)\n",
    "    acc_sadness = model_rf(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_rf.append(acc_sadness)\n",
    "    print(\"Random Forest: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while feature number of trees is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gradient Boosting: dev set RMSE for anger is 0.0096 while number of stages is 10.00\n",
      "Gradient Boosting: dev set RMSE for fear is 0.0086 while number of stages is 10.00\n",
      "Gradient Boosting: dev set RMSE for joy is 0.0102 while number of stages is 10.00\n",
      "Gradient Boosting: dev set RMSE for sadness is 0.0091 while number of stages is 10.00\n",
      "Gradient Boosting: dev set RMSE for anger is 0.0091 while number of stages is 60.00\n",
      "Gradient Boosting: dev set RMSE for fear is 0.0082 while number of stages is 60.00\n",
      "Gradient Boosting: dev set RMSE for joy is 0.0100 while number of stages is 60.00\n",
      "Gradient Boosting: dev set RMSE for sadness is 0.0086 while number of stages is 60.00\n",
      "Gradient Boosting: dev set RMSE for anger is 0.0090 while number of stages is 110.00\n",
      "Gradient Boosting: dev set RMSE for fear is 0.0081 while number of stages is 110.00\n",
      "Gradient Boosting: dev set RMSE for joy is 0.0099 while number of stages is 110.00\n",
      "Gradient Boosting: dev set RMSE for sadness is 0.0086 while number of stages is 110.00\n",
      "Gradient Boosting: dev set RMSE for anger is 0.0089 while number of stages is 160.00\n",
      "Gradient Boosting: dev set RMSE for fear is 0.0081 while number of stages is 160.00\n",
      "Gradient Boosting: dev set RMSE for joy is 0.0099 while number of stages is 160.00\n",
      "Gradient Boosting: dev set RMSE for sadness is 0.0086 while number of stages is 160.00\n",
      "Gradient Boosting: dev set RMSE for anger is 0.0089 while number of stages is 210.00\n",
      "Gradient Boosting: dev set RMSE for fear is 0.0080 while number of stages is 210.00\n",
      "Gradient Boosting: dev set RMSE for joy is 0.0099 while number of stages is 210.00\n",
      "Gradient Boosting: dev set RMSE for sadness is 0.0086 while number of stages is 210.00\n",
      "Gradient Boosting: dev set RMSE for anger is 0.0089 while number of stages is 260.00\n",
      "Gradient Boosting: dev set RMSE for fear is 0.0080 while number of stages is 260.00\n",
      "Gradient Boosting: dev set RMSE for joy is 0.0098 while number of stages is 260.00\n",
      "Gradient Boosting: dev set RMSE for sadness is 0.0086 while number of stages is 260.00\n"
     ]
    }
   ],
   "source": [
    "#Gradient Boosting Regressor\n",
    "\n",
    "def model_gb(trainX, trainY, devX, devY, c):\n",
    "    model = GradientBoostingRegressor(n_estimators = c)\n",
    "    model.fit(trainX, trainY)\n",
    "    train_predict = model.predict(devX)\n",
    "    train_acc = np.linalg.norm(train_predict - devY) / len(train_predict)  \n",
    "    return train_acc\n",
    " \n",
    "stages_number = np.arange(10, 261, 50)\n",
    "accuracy_anger_gb = []\n",
    "accuracy_fear_gb = []\n",
    "accuracy_joy_gb = []\n",
    "accuracy_sadness_gb = []\n",
    "for c in stages_number:\n",
    "    acc_anger = model_gb(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "    accuracy_anger_gb.append(acc_anger)\n",
    "    print(\"Gradient Boosting: dev set RMSE for anger is %.4f\" % acc_anger, \"while number of stages is %.2f\" % c)\n",
    "    acc_fear = model_gb(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "    accuracy_fear_gb.append(acc_fear)\n",
    "    print(\"Gradient Boosting: dev set RMSE for fear is %.4f\" % acc_fear, \"while number of stages is %.2f\" % c)\n",
    "    acc_joy = model_gb(train_joy, joy_indexes, dev_joy, joy_indexes_dev, c)\n",
    "    accuracy_joy_gb.append(acc_joy)\n",
    "    print(\"Gradient Boosting: dev set RMSE for joy is %.4f\" % acc_joy, \"while number of stages is %.2f\" % c)\n",
    "    acc_sadness = model_gb(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "    accuracy_sadness_gb.append(acc_sadness)\n",
    "    print(\"Gradient Boosting: dev set RMSE for sadness is %.4f\" % acc_sadness, \"while number of stages is %.2f\" % c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The best RMSE for anger is 0.0088\n",
      "The best RMSE for fear is 0.0079\n",
      "The best RMSE for joy is 0.0097\n",
      "The best RMSE for sadness is 0.0084\n"
     ]
    }
   ],
   "source": [
    "#The best models\n",
    "\n",
    "#anger \n",
    "c = 0.01\n",
    "acc_anger = model_svm_lin(train_anger, anger_indexes, dev_anger, anger_indexes_dev, c)\n",
    "print(\"The best RMSE for anger is %.4f\" % acc_anger)\n",
    "#fear\n",
    "acc_fear = model_svm_lin(train_fear, fear_indexes, dev_fear, fear_indexes_dev, c)\n",
    "print(\"The best RMSE for fear is %.4f\" % acc_fear)\n",
    "#joy\n",
    "s = 1\n",
    "acc_joy = model_svm_poly(train_joy, joy_indexes, dev_joy, joy_indexes_dev, s)\n",
    "print(\"The best RMSE for joy is %.4f\" % acc_joy)\n",
    "#sadness\n",
    "acc_sadness = model_svm_lin(train_sadness, sadness_indexes, dev_sadness, sadness_indexes_dev, c)\n",
    "print(\"The best RMSE for sadness is %.4f\" % acc_sadness)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "﻿Our #NLP course is #awesome, we all #love it! Interesting tasks, perfect explanations and high challenges! That was #excellent! 😊\n",
      "\n",
      "I hate this stupid #morning !\n",
      "\n",
      "Too scare not to pass my #course s...\n",
      "\n",
      "I'm so lonely, really miss my mommy :(\n",
      "\n",
      "Seriously @Comcast, why must my wireless signal go out intermittently for hours/days every couple of weeks? I am so over this. #evil\n",
      "\n",
      "Didn't think I could be any more disappointed with my country's criminal justice system. I was wrong. #EricGarner\n",
      "\n",
      "Happy birthday to the coolest, the most adorable people in the World - my lovely boyfriend! Hope you will have an awesome day! 😍😘 \n",
      "[['\\ufeff', 'Our', 'NLP', 'course', 'awesome', 'love', 'Interesting', 'tasks', 'perfect', 'explanations', 'high', 'challenges', 'That', 'excellent', '😊'], ['I', 'hate', 'stupid', 'morning'], ['Too', 'scare', 'pass', 'course'], [\"I'm\", 'lonely', 'really', 'miss', 'mommy', ':('], ['Seriously', 'must', 'wireless', 'signal', 'go', 'intermittently', 'hours', 'days', 'every', 'couple', 'weeks', 'I', 'evil'], [\"Didn't\", 'think', 'I', 'could', 'disappointed', \"country's\", 'criminal', 'justice', 'system', 'I', 'wrong', 'EricGarner'], ['Happy', 'birthday', 'coolest', 'adorable', 'people', 'World', 'lovely', 'boyfriend', 'Hope', 'awesome', 'day', '😍', '😘']]\n",
      "Intensity of anger:  [ 0.45374879  0.62222645  0.32948608  0.49508109  0.58254224  0.55459265\n",
      "  0.34290958]\n",
      "Intensity of fear:  [ 0.25698106  0.67921924  0.44673096  0.66875685  0.56925059  0.63454808\n",
      "  0.4396304 ]\n",
      "Intensity of joy:  [ 0.60515202  0.48683256  0.42594276  0.53773462  0.45926113  0.37930136\n",
      "  0.76617036]\n",
      "Intensity of sadness:  [ 0.39357734  0.6273273   0.48537257  0.5696813   0.53943233  0.56700746\n",
      "  0.44879174]\n"
     ]
    }
   ],
   "source": [
    "#Testing new tweet\n",
    "\n",
    "#Read data\n",
    "tweets = {}\n",
    "with open(\"test.txt\", encoding=\"utf8\") as f:\n",
    "    for line in f:\n",
    "        key = line\n",
    "        print (line)\n",
    "        tweets[key] = len(line)\n",
    "\n",
    "#Made tokens\n",
    "tweets_tokens = []\n",
    "tweets_ind = []\n",
    "filter_dict(tweets, tweets_tokens, tweets_ind)\n",
    "print(tweets_tokens)\n",
    "\n",
    "#Features extracting\n",
    "tweets_dicts = get_feat(tweets_tokens)\n",
    "\n",
    "#Emotions intensity\n",
    "#anger\n",
    "model_anger = SVR(kernel='linear', C = 0.01, degree = 2)\n",
    "model_anger.fit(train_anger, anger_indexes)\n",
    "test_predict_anger = model_anger.predict(tweets_dicts)\n",
    "print (\"Intensity of anger: \", test_predict_anger)\n",
    "#fear\n",
    "model_fear = SVR(kernel='linear', C = 0.01, degree = 2)\n",
    "model_fear.fit(train_fear, fear_indexes)\n",
    "test_predict_fear = model_fear.predict(tweets_dicts)\n",
    "print (\"Intensity of fear: \", test_predict_fear)\n",
    "#joy\n",
    "model_joy = SVR(kernel='poly', C = 1, degree = 2)\n",
    "model_joy.fit(train_joy, joy_indexes)\n",
    "test_predict_joy = model_joy.predict(tweets_dicts)\n",
    "print (\"Intensity of joy: \", test_predict_joy)\n",
    "#sadness\n",
    "model_sadness = SVR(kernel='linear', C = 0.01, degree = 2)\n",
    "model_sadness.fit(train_sadness, sadness_indexes)\n",
    "test_predict_sadness = model_sadness.predict(tweets_dicts)\n",
    "print (\"Intensity of sadness: \", test_predict_sadness)"
   ]
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}