Capstone_part4-Copy3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd  \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('fivethirtyeight')\n",
    "\n",
    "%matplotlib inline\n",
    "%config InlineBackend.figure_format = 'retina'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>awww that bummer you shoulda got david carr of...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>is upset that he can not update his facebook b...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dived many times for the ball managed to save ...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>my whole body feels itchy and like its on fire</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>no it not behaving at all mad why am here beca...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text  target\n",
       "0  awww that bummer you shoulda got david carr of...       0\n",
       "1  is upset that he can not update his facebook b...       0\n",
       "2  dived many times for the ball managed to save ...       0\n",
       "3     my whole body feels itchy and like its on fire       0\n",
       "4  no it not behaving at all mad why am here beca...       0"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "csv = 'clean_tweet.csv'\n",
    "my_df = pd.read_csv(csv,index_col=0)\n",
    "my_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1596019 entries, 0 to 1596018\n",
      "Data columns (total 2 columns):\n",
      "text      1596019 non-null object\n",
      "target    1596019 non-null int64\n",
      "dtypes: int64(1), object(1)\n",
      "memory usage: 24.4+ MB\n"
     ]
    }
   ],
   "source": [
    "my_df.dropna(inplace=True)\n",
    "my_df.reset_index(drop=True,inplace=True)\n",
    "my_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x = my_df.text\n",
    "y = my_df.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "SEED = 2000\n",
    "x_train, x_validation_and_test, y_train, y_validation_and_test = train_test_split(x, y, test_size=.02, random_state=SEED)\n",
    "x_validation, x_test, y_validation, y_test = train_test_split(x_validation_and_test, y_validation_and_test, test_size=.5, random_state=SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.pipeline import Pipeline\n",
    "from time import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def accuracy_summary(pipeline, x_train, y_train, x_test, y_test):\n",
    "    if len(x_test[y_test == 0]) / (len(x_test)*1.) > 0.5:\n",
    "        null_accuracy = len(x_test[y_test == 0]) / (len(x_test)*1.)\n",
    "    else:\n",
    "        null_accuracy = 1. - (len(x_test[y_test == 0]) / (len(x_test)*1.))\n",
    "    t0 = time()\n",
    "    sentiment_fit = pipeline.fit(x_train, y_train)\n",
    "    y_pred = sentiment_fit.predict(x_test)\n",
    "    train_test_time = time() - t0\n",
    "    accuracy = accuracy_score(y_test, y_pred)\n",
    "    print \"null accuracy: {0:.2f}%\".format(null_accuracy*100)\n",
    "    print \"accuracy score: {0:.2f}%\".format(accuracy*100)\n",
    "    if accuracy > null_accuracy:\n",
    "        print \"model is {0:.2f}% more accurate than null accuracy\".format((accuracy-null_accuracy)*100)\n",
    "    elif accuracy == null_accuracy:\n",
    "        print \"model has the same accuracy with the null accuracy\"\n",
    "    else:\n",
    "        print \"model is {0:.2f}% less accurate than null accuracy\".format((null_accuracy-accuracy)*100)\n",
    "    print \"train and test time: {0:.2f}s\".format(train_test_time)\n",
    "    print \"-\"*80\n",
    "    return accuracy, train_test_time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cvec = CountVectorizer()\n",
    "lr = LogisticRegression()\n",
    "n_features = np.arange(10000,100001,10000)\n",
    "\n",
    "def nfeature_accuracy_checker(vectorizer=cvec, n_features=n_features, stop_words=None, ngram_range=(1, 1), classifier=lr):\n",
    "    result = []\n",
    "    print (classifier)\n",
    "    print \"\\n\"\n",
    "    for n in n_features:\n",
    "        vectorizer.set_params(stop_words=stop_words, max_features=n, ngram_range=ngram_range)\n",
    "        checker_pipeline = Pipeline([\n",
    "            ('vectorizer', vectorizer),\n",
    "            ('classifier', classifier)\n",
    "        ])\n",
    "        print \"Validation result for {} features\".format(n)\n",
    "        nfeature_accuracy,tt_time = accuracy_summary(checker_pipeline, x_train, y_train, x_validation, y_validation)\n",
    "        result.append((n,nfeature_accuracy,tt_time))\n",
    "    return result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### TFIDF Vectorizer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TF-IDF is another way to convert textual data to a numeric form and is short for Term Frequency-Inverse Document Frequency. The vector value it yields is the product of these two terms; TF and IDF."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's first look at Term Frequency. We have already looked at term frequency above with count vectorizer, but this time, we need one more step to calculate the relative frequency. Let's say we have two documents in total as below.\n",
    "\n",
    "1. I love dogs\n",
    "2. I hate dogs and knitting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Relative term frequency is calculated for each term within each document as below.\n",
    "\n",
    "$${TF(t,d)} = \\frac {number\\ of\\ times\\ term(t)\\ appears\\ in\\ document(d)}{total\\ number\\ of\\ terms\\ in\\ document(d)}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For example, if we calculate relative term frequency for 'I' in both document 1 and document 2, it will be as below.\n",
    "\n",
    "$${TF('I',d1)} = \\frac {1}{3} \\approx {0.33}$$\n",
    "\n",
    "$${TF('I',d2)} = \\frac {1}{5} = {0.2}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we need to get Inverse Document Frequency, which measures how important a word is to differentiate each document by following the calculation as below.\n",
    "\n",
    "$${IDF(t,D)} = \\log \\Big(\\frac {total\\ number\\ of\\ documents(D)}{number\\ of\\ documents\\ with\\ the\\ term(t)\\ in\\ it}\\Big)$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we calculate inverse document frequency for 'I',\n",
    "\n",
    "$${IDF('I',D)} = \\log \\Big(\\frac {2}{2}\\Big) = {0}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once we have the values for TF and IDF, now we can calculate TFIDF as below.\n",
    "\n",
    "$${TFIDF(t,d,D)} = {TF(t,d)}\\cdot{IDF(t,D)}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Following the case of our example, TFIDF for term 'I' in both documents will be as below.\n",
    "\n",
    "$${TFIDF('I',d1,D)} = {TF('I',d1)}\\cdot{IDF('I',D)} = {0.33}\\times{0} = {0}$$\n",
    "\n",
    "$${TFIDF('I',d2,D)} = {TF('I',d2)}\\cdot{IDF('I',D)} = {0.2}\\times{0} = {0}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, the term 'I' appeared equally in both documents, and the TFIDF score is 0, which means the term is not really informative in differentiating documents. The rest is same as count vectorizer, TFIDF vectorizer will calculate these scores for terms in documents, and convert textual data into a numeric form."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "tvec = TfidfVectorizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RESULT FOR UNIGRAM WITH STOP WORDS (Tfidf)\n",
      "\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "\n",
      "\n",
      "Validation result for 10000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.12%\n",
      "model is 29.72% more accurate than null accuracy\n",
      "train and test time: 58.64s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 20000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.21%\n",
      "model is 29.81% more accurate than null accuracy\n",
      "train and test time: 65.26s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 30000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.36%\n",
      "model is 29.96% more accurate than null accuracy\n",
      "train and test time: 63.47s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 40000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.35%\n",
      "model is 29.95% more accurate than null accuracy\n",
      "train and test time: 64.60s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 50000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.43%\n",
      "model is 30.03% more accurate than null accuracy\n",
      "train and test time: 65.32s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 60000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.45%\n",
      "model is 30.05% more accurate than null accuracy\n",
      "train and test time: 60.28s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 70000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.47%\n",
      "model is 30.07% more accurate than null accuracy\n",
      "train and test time: 67.03s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 80000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.42%\n",
      "model is 30.02% more accurate than null accuracy\n",
      "train and test time: 68.01s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 90000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.46%\n",
      "model is 30.06% more accurate than null accuracy\n",
      "train and test time: 64.28s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 100000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.45%\n",
      "model is 30.05% more accurate than null accuracy\n",
      "train and test time: 67.47s\n",
      "--------------------------------------------------------------------------------\n",
      "CPU times: user 10min 24s, sys: 34.1 s, total: 10min 58s\n",
      "Wall time: 10min 44s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print \"RESULT FOR UNIGRAM WITH STOP WORDS (Tfidf)\\n\"\n",
    "feature_result_ugt = nfeature_accuracy_checker(vectorizer=tvec)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RESULT FOR BIGRAM WITH STOP WORDS (Tfidf)\n",
      "\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "\n",
      "\n",
      "Validation result for 10000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 81.26%\n",
      "model is 30.86% more accurate than null accuracy\n",
      "train and test time: 134.08s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 20000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 81.84%\n",
      "model is 31.43% more accurate than null accuracy\n",
      "train and test time: 141.47s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 30000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.17%\n",
      "model is 31.77% more accurate than null accuracy\n",
      "train and test time: 139.81s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 40000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.17%\n",
      "model is 31.77% more accurate than null accuracy\n",
      "train and test time: 139.30s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 50000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.47%\n",
      "model is 32.07% more accurate than null accuracy\n",
      "train and test time: 141.54s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 60000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.53%\n",
      "model is 32.12% more accurate than null accuracy\n",
      "train and test time: 141.64s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 70000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.56%\n",
      "model is 32.16% more accurate than null accuracy\n",
      "train and test time: 142.88s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 80000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.58%\n",
      "model is 32.18% more accurate than null accuracy\n",
      "train and test time: 145.13s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 90000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.68%\n",
      "model is 32.27% more accurate than null accuracy\n",
      "train and test time: 139.80s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 100000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.77%\n",
      "model is 32.37% more accurate than null accuracy\n",
      "train and test time: 140.77s\n",
      "--------------------------------------------------------------------------------\n",
      "CPU times: user 22min 43s, sys: 1min 7s, total: 23min 51s\n",
      "Wall time: 23min 26s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print \"RESULT FOR BIGRAM WITH STOP WORDS (Tfidf)\\n\"\n",
    "feature_result_bgt = nfeature_accuracy_checker(vectorizer=tvec,ngram_range=(1, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RESULT FOR TRIGRAM WITH STOP WORDS (Tfidf)\n",
      "\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "\n",
      "\n",
      "Validation result for 10000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 81.24%\n",
      "model is 30.84% more accurate than null accuracy\n",
      "train and test time: 285.57s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 20000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 81.99%\n",
      "model is 31.59% more accurate than null accuracy\n",
      "train and test time: 297.29s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 30000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.16%\n",
      "model is 31.75% more accurate than null accuracy\n",
      "train and test time: 300.72s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 40000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.44%\n",
      "model is 32.04% more accurate than null accuracy\n",
      "train and test time: 300.80s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 50000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.63%\n",
      "model is 32.22% more accurate than null accuracy\n",
      "train and test time: 304.85s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 60000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.74%\n",
      "model is 32.34% more accurate than null accuracy\n",
      "train and test time: 303.27s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 70000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.77%\n",
      "model is 32.37% more accurate than null accuracy\n",
      "train and test time: 306.02s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 80000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.88%\n",
      "model is 32.47% more accurate than null accuracy\n",
      "train and test time: 307.10s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 90000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.89%\n",
      "model is 32.49% more accurate than null accuracy\n",
      "train and test time: 310.60s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for 100000 features\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.92%\n",
      "model is 32.52% more accurate than null accuracy\n",
      "train and test time: 310.84s\n",
      "--------------------------------------------------------------------------------\n",
      "CPU times: user 46min 49s, sys: 3min 41s, total: 50min 30s\n",
      "Wall time: 50min 27s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print \"RESULT FOR TRIGRAM WITH STOP WORDS (Tfidf)\\n\"\n",
    "feature_result_tgt = nfeature_accuracy_checker(vectorizer=tvec,ngram_range=(1, 3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It seems like TFIDF vectorizer is yielding better results when fed to logistic regression. Let's plot the results from count vectorizer together with TFIDF vectorizer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2055601d0>"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABGcAAAMFCAYAAADZTJu3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x133d9bf90>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 386,
       "width": 563
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "nfeatures_plot_tgt = pd.DataFrame(feature_result_tgt,columns=['nfeatures','validation_accuracy','train_test_time'])\n",
    "nfeatures_plot_bgt = pd.DataFrame(feature_result_bgt,columns=['nfeatures','validation_accuracy','train_test_time'])\n",
    "nfeatures_plot_ugt = pd.DataFrame(feature_result_ugt,columns=['nfeatures','validation_accuracy','train_test_time'])\n",
    "\n",
    "plt.figure(figsize=(8,6))\n",
    "plt.plot(nfeatures_plot_tgt.nfeatures, nfeatures_plot_tgt.validation_accuracy,label='trigram tfidf vectorizer',color='royalblue')\n",
    "plt.plot(nfeatures_plot_tg.nfeatures, nfeatures_plot_tg.validation_accuracy,label='trigram count vectorizer',linestyle=':', color='royalblue')\n",
    "plt.plot(nfeatures_plot_bgt.nfeatures, nfeatures_plot_bgt.validation_accuracy,label='bigram tfidf vectorizer',color='orangered')\n",
    "plt.plot(nfeatures_plot_bg.nfeatures, nfeatures_plot_bg.validation_accuracy,label='bigram count vectorizer',linestyle=':',color='orangered')\n",
    "plt.plot(nfeatures_plot_ugt.nfeatures, nfeatures_plot_ugt.validation_accuracy, label='unigram tfidf vectorizer',color='gold')\n",
    "plt.plot(nfeatures_plot_ug.nfeatures, nfeatures_plot_ug.validation_accuracy, label='unigram count vectorizer',linestyle=':',color='gold')\n",
    "plt.title(\"N-gram(1~3) test result : Accuracy\")\n",
    "plt.xlabel(\"Number of features\")\n",
    "plt.ylabel(\"Validation set accuracy\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From above chart, we can see including bigram and trigram boost the model performance both in count vectorizer and TFIDF vectorizer. And for every case of unigram to trigram, TFIDF yields better results than count vectorizer."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Algorithms Comparison"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The best result I can get with logistic regression was by using TFIDF vectorizer of 100,000 features including up to trigram. With this I will first fit various different models and compare their validation results, then I will build an ensemble (voting) classifier with top 5 models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I haven't included some of computationally expensive models, such as KNN, random forest, considering the size of data and the scalability of models. And the fine-tuning of models will come after I try some other different vectorisation of textual data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n",
    "from sklearn.naive_bayes import MultinomialNB, BernoulliNB\n",
    "from sklearn.linear_model import RidgeClassifier\n",
    "from sklearn.linear_model import PassiveAggressiveClassifier\n",
    "from sklearn.linear_model import Perceptron\n",
    "from sklearn.neighbors import NearestCentroid\n",
    "from sklearn.feature_selection import SelectFromModel\n",
    "\n",
    "names = [\"Logistic Regression\", \"Linear SVC\", \"LinearSVC with L1-based feature selection\",\"Multinomial NB\", \n",
    "         \"Bernoulli NB\", \"Ridge Classifier\", \"AdaBoost\", \"Perceptron\",\"Passive-Aggresive\", \"Nearest Centroid\"]\n",
    "classifiers = [\n",
    "    LogisticRegression(),\n",
    "    LinearSVC(),\n",
    "    Pipeline([\n",
    "  ('feature_selection', SelectFromModel(LinearSVC(penalty=\"l1\", dual=False))),\n",
    "  ('classification', LinearSVC(penalty=\"l2\"))]),\n",
    "    MultinomialNB(),\n",
    "    BernoulliNB(),\n",
    "    RidgeClassifier(),\n",
    "    AdaBoostClassifier(),\n",
    "    Perceptron(),\n",
    "    PassiveAggressiveClassifier(),\n",
    "    NearestCentroid()\n",
    "    ]\n",
    "zipped_clf = zip(names,classifiers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tvec = TfidfVectorizer()\n",
    "def classifier_comparator(vectorizer=tvec, n_features=10000, stop_words=None, ngram_range=(1, 1), classifier=zipped_clf):\n",
    "    result = []\n",
    "    vectorizer.set_params(stop_words=stop_words, max_features=n_features, ngram_range=ngram_range)\n",
    "    for n,c in classifier:\n",
    "        checker_pipeline = Pipeline([\n",
    "            ('vectorizer', vectorizer),\n",
    "            ('classifier', c)\n",
    "        ])\n",
    "        print \"Validation result for {}\".format(n)\n",
    "        print c\n",
    "        clf_accuracy,tt_time = accuracy_summary(checker_pipeline, x_train, y_train, x_validation, y_validation)\n",
    "        result.append((n,clf_accuracy,tt_time))\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation result for Logistic Regression\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.92%\n",
      "model is 32.52% more accurate than null accuracy\n",
      "train and test time: 325.26s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Linear SVC\n",
      "LinearSVC(C=1.0, class_weight=None, dual=True, fit_intercept=True,\n",
      "     intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
      "     multi_class='ovr', penalty='l2', random_state=None, tol=0.0001,\n",
      "     verbose=0)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.22%\n",
      "model is 31.82% more accurate than null accuracy\n",
      "train and test time: 358.16s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for LinearSVC with L1-based feature selection\n",
      "Pipeline(memory=None,\n",
      "     steps=[('feature_selection', SelectFromModel(estimator=LinearSVC(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "     intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
      "     multi_class='ovr', penalty='l1', random_state=None, tol=0.0001,\n",
      "     verbose=0),\n",
      "        norm_order=1, prefit...ax_iter=1000,\n",
      "     multi_class='ovr', penalty='l2', random_state=None, tol=0.0001,\n",
      "     verbose=0))])\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.18%\n",
      "model is 31.78% more accurate than null accuracy\n",
      "train and test time: 489.43s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Multinomial NB\n",
      "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.09%\n",
      "model is 29.69% more accurate than null accuracy\n",
      "train and test time: 316.77s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Bernoulli NB\n",
      "BernoulliNB(alpha=1.0, binarize=0.0, class_prior=None, fit_prior=True)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 79.20%\n",
      "model is 28.80% more accurate than null accuracy\n",
      "train and test time: 320.55s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Ridge Classifier\n",
      "RidgeClassifier(alpha=1.0, class_weight=None, copy_X=True, fit_intercept=True,\n",
      "        max_iter=None, normalize=False, random_state=None, solver='auto',\n",
      "        tol=0.001)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.14%\n",
      "model is 31.74% more accurate than null accuracy\n",
      "train and test time: 379.79s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for AdaBoost\n",
      "AdaBoostClassifier(algorithm='SAMME.R', base_estimator=None,\n",
      "          learning_rate=1.0, n_estimators=50, random_state=None)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 69.47%\n",
      "model is 19.07% more accurate than null accuracy\n",
      "train and test time: 1194.67s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Perceptron\n",
      "Perceptron(alpha=0.0001, class_weight=None, eta0=1.0, fit_intercept=True,\n",
      "      max_iter=5, n_iter=None, n_jobs=1, penalty=None, random_state=0,\n",
      "      shuffle=True, tol=None, verbose=0, warm_start=False)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 75.57%\n",
      "model is 25.17% more accurate than null accuracy\n",
      "train and test time: 331.08s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Passive-Aggresive\n",
      "PassiveAggressiveClassifier(C=1.0, average=False, class_weight=None,\n",
      "              fit_intercept=True, loss='hinge', max_iter=5, n_iter=None,\n",
      "              n_jobs=1, random_state=None, shuffle=True, tol=None,\n",
      "              verbose=0, warm_start=False)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 79.47%\n",
      "model is 29.07% more accurate than null accuracy\n",
      "train and test time: 328.61s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Nearest Centroid\n",
      "NearestCentroid(metric='euclidean', shrink_threshold=None)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 72.74%\n",
      "model is 22.34% more accurate than null accuracy\n",
      "train and test time: 336.03s\n",
      "--------------------------------------------------------------------------------\n",
      "CPU times: user 1h 6min 23s, sys: 4min 21s, total: 1h 10min 44s\n",
      "Wall time: 1h 13min\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "trigram_result = classifier_comparator(n_features=100000,ngram_range=(1,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation result for Logistic Regression\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.92%\n",
      "model is 32.52% more accurate than null accuracy\n",
      "train and test time: 319.69s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Linear SVC\n",
      "LinearSVC(C=1.0, class_weight=None, dual=True, fit_intercept=True,\n",
      "     intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
      "     multi_class='ovr', penalty='l2', random_state=None, tol=0.0001,\n",
      "     verbose=0)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.22%\n",
      "model is 31.82% more accurate than null accuracy\n",
      "train and test time: 297.98s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Multinomial NB\n",
      "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 80.09%\n",
      "model is 29.69% more accurate than null accuracy\n",
      "train and test time: 258.20s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Ridge Classifier\n",
      "RidgeClassifier(alpha=1.0, class_weight=None, copy_X=True, fit_intercept=True,\n",
      "        max_iter=None, normalize=False, random_state=None, solver='auto',\n",
      "        tol=0.001)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.15%\n",
      "model is 31.75% more accurate than null accuracy\n",
      "train and test time: 315.52s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Passive Aggresive Classifier\n",
      "PassiveAggressiveClassifier(C=1.0, average=False, class_weight=None,\n",
      "              fit_intercept=True, loss='hinge', max_iter=5, n_iter=None,\n",
      "              n_jobs=1, random_state=None, shuffle=True, tol=None,\n",
      "              verbose=0, warm_start=False)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 79.66%\n",
      "model is 29.25% more accurate than null accuracy\n",
      "train and test time: 262.14s\n",
      "--------------------------------------------------------------------------------\n",
      "Validation result for Ensemble\n",
      "VotingClassifier(estimators=[('lr', LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)), ('svc', Linear...   n_jobs=1, random_state=None, shuffle=True, tol=None,\n",
      "              verbose=0, warm_start=False))],\n",
      "         flatten_transform=None, n_jobs=1, voting='hard', weights=None)\n",
      "null accuracy: 50.40%\n",
      "accuracy score: 82.46%\n",
      "model is 32.06% more accurate than null accuracy\n",
      "train and test time: 388.05s\n",
      "--------------------------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import VotingClassifier\n",
    "\n",
    "clf1 = LogisticRegression()\n",
    "clf2 = LinearSVC()\n",
    "clf3 = MultinomialNB()\n",
    "clf4 = RidgeClassifier()\n",
    "clf5 = PassiveAggressiveClassifier()\n",
    "\n",
    "eclf = VotingClassifier(estimators=[('lr', clf1), ('svc', clf2), ('mnb', clf3), ('rcs', clf4), ('pac', clf5)], voting='hard')\n",
    "\n",
    "for clf, label in zip([clf1, clf2, clf3, clf4, clf5, eclf], ['Logistic Regression', 'Linear SVC', 'Multinomial NB', 'Ridge Classifier', 'Passive Aggresive Classifier', 'Ensemble']):\n",
    "    checker_pipeline = Pipeline([\n",
    "            ('vectorizer', TfidfVectorizer(max_features=100000,ngram_range=(1, 3))),\n",
    "            ('classifier', clf)\n",
    "        ])\n",
    "    print \"Validation result for {}\".format(label)\n",
    "    print clf\n",
    "    clf_accuracy,tt_time = accuracy_summary(checker_pipeline, x_train, y_train, x_validation, y_validation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It seems like the voting classifier does no better than the simple logistic regression model. Thus later part, I will try to finetune logistic regression model. But before that, I would like to try another method of sentiment classification."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Lexical approach"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What I have demonstrated above are machine learning approaches to text classification problem, which tries to solve the problem by training classifiers on the labelled data set. Another famous approach to sentiment analysis task is a lexical approach. \"In the lexical approach the definition of sentiment is based on the analysis of individual words and/or phrases; emotional dictionaries are often used: emotional lexical items from the dictionary are searched in the text, their sentiment weights are calculated, and some aggregated weight function is applied.\"\n",
    "http://www.dialog-21.ru/media/1226/blinovpd.pdf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the previous part, I have calculated harmonic mean of \"positive rate CDF\" and \"positive frequency percent CDF\", and these have given me a good representation of positive and negative terms in the corpora. If it successfully filters which terms are important to each class, then this can also be used for prediction in a lexical manner."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So I decided to make a simple predictor, which make use of the harmonic mean value I calculated. Below I go through the term frequency calculation, and the steps to get 'pos_normcdf_hmean', but this time I calculated term frequency only from the train set. (* Since I learned that I don't need to transform sparse matrix to dense matrix for term frequency calculation, I computed the frequency directly from sparse matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CountVectorizer(analyzer=u'word', binary=False, decode_error=u'strict',\n",
       "        dtype=<type 'numpy.int64'>, encoding=u'utf-8', input=u'content',\n",
       "        lowercase=True, max_df=1.0, max_features=10000, min_df=1,\n",
       "        ngram_range=(1, 1), preprocessor=None, stop_words=None,\n",
       "        strip_accents=None, token_pattern=u'(?u)\\\\b\\\\w\\\\w+\\\\b',\n",
       "        tokenizer=None, vocabulary=None)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "cvec = CountVectorizer(max_features=10000)\n",
    "cvec.fit(x_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "neg_train = x_train[y_train == 0]\n",
    "pos_train = x_train[y_train == 1]\n",
    "neg_doc_matrix = cvec.transform(neg_train)\n",
    "pos_doc_matrix = cvec.transform(pos_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 53.5 ms, sys: 5.62 ms, total: 59.2 ms\n",
      "Wall time: 57.6 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "neg_tf = np.sum(neg_doc_matrix,axis=0)\n",
    "pos_tf = np.sum(pos_doc_matrix,axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>negative</th>\n",
       "      <th>positive</th>\n",
       "      <th>total</th>\n",
       "      <th>pos_rate</th>\n",
       "      <th>pos_freq_pct</th>\n",
       "      <th>pos_rate_normcdf</th>\n",
       "      <th>pos_freq_pct_normcdf</th>\n",
       "      <th>pos_normcdf_hmean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>thanks</th>\n",
       "      <td>5646</td>\n",
       "      <td>33697</td>\n",
       "      <td>39343</td>\n",
       "      <td>0.856493</td>\n",
       "      <td>0.004060</td>\n",
       "      <td>0.988672</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.994304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thank</th>\n",
       "      <td>2234</td>\n",
       "      <td>15428</td>\n",
       "      <td>17662</td>\n",
       "      <td>0.873514</td>\n",
       "      <td>0.001859</td>\n",
       "      <td>0.991521</td>\n",
       "      <td>0.990691</td>\n",
       "      <td>0.991106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>awesome</th>\n",
       "      <td>3735</td>\n",
       "      <td>14189</td>\n",
       "      <td>17924</td>\n",
       "      <td>0.791620</td>\n",
       "      <td>0.001709</td>\n",
       "      <td>0.968984</td>\n",
       "      <td>0.984355</td>\n",
       "      <td>0.976609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>happy</th>\n",
       "      <td>6371</td>\n",
       "      <td>20123</td>\n",
       "      <td>26494</td>\n",
       "      <td>0.759530</td>\n",
       "      <td>0.002424</td>\n",
       "      <td>0.951714</td>\n",
       "      <td>0.999064</td>\n",
       "      <td>0.974815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>great</th>\n",
       "      <td>8026</td>\n",
       "      <td>24819</td>\n",
       "      <td>32845</td>\n",
       "      <td>0.755640</td>\n",
       "      <td>0.002990</td>\n",
       "      <td>0.949178</td>\n",
       "      <td>0.999945</td>\n",
       "      <td>0.973900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>love</th>\n",
       "      <td>16657</td>\n",
       "      <td>46737</td>\n",
       "      <td>63394</td>\n",
       "      <td>0.737246</td>\n",
       "      <td>0.005631</td>\n",
       "      <td>0.935719</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.966792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hey</th>\n",
       "      <td>4945</td>\n",
       "      <td>13769</td>\n",
       "      <td>18714</td>\n",
       "      <td>0.735759</td>\n",
       "      <td>0.001659</td>\n",
       "      <td>0.934520</td>\n",
       "      <td>0.981496</td>\n",
       "      <td>0.957432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nice</th>\n",
       "      <td>6468</td>\n",
       "      <td>16632</td>\n",
       "      <td>23100</td>\n",
       "      <td>0.720000</td>\n",
       "      <td>0.002004</td>\n",
       "      <td>0.920725</td>\n",
       "      <td>0.994570</td>\n",
       "      <td>0.956224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>yay</th>\n",
       "      <td>3106</td>\n",
       "      <td>10275</td>\n",
       "      <td>13381</td>\n",
       "      <td>0.767880</td>\n",
       "      <td>0.001238</td>\n",
       "      <td>0.956816</td>\n",
       "      <td>0.936053</td>\n",
       "      <td>0.946321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>haha</th>\n",
       "      <td>9633</td>\n",
       "      <td>21261</td>\n",
       "      <td>30894</td>\n",
       "      <td>0.688192</td>\n",
       "      <td>0.002561</td>\n",
       "      <td>0.886360</td>\n",
       "      <td>0.999505</td>\n",
       "      <td>0.939538</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         negative  positive  total  pos_rate  pos_freq_pct  pos_rate_normcdf  \\\n",
       "thanks       5646     33697  39343  0.856493      0.004060          0.988672   \n",
       "thank        2234     15428  17662  0.873514      0.001859          0.991521   \n",
       "awesome      3735     14189  17924  0.791620      0.001709          0.968984   \n",
       "happy        6371     20123  26494  0.759530      0.002424          0.951714   \n",
       "great        8026     24819  32845  0.755640      0.002990          0.949178   \n",
       "love        16657     46737  63394  0.737246      0.005631          0.935719   \n",
       "hey          4945     13769  18714  0.735759      0.001659          0.934520   \n",
       "nice         6468     16632  23100  0.720000      0.002004          0.920725   \n",
       "yay          3106     10275  13381  0.767880      0.001238          0.956816   \n",
       "haha         9633     21261  30894  0.688192      0.002561          0.886360   \n",
       "\n",
       "         pos_freq_pct_normcdf  pos_normcdf_hmean  \n",
       "thanks               1.000000           0.994304  \n",
       "thank                0.990691           0.991106  \n",
       "awesome              0.984355           0.976609  \n",
       "happy                0.999064           0.974815  \n",
       "great                0.999945           0.973900  \n",
       "love                 1.000000           0.966792  \n",
       "hey                  0.981496           0.957432  \n",
       "nice                 0.994570           0.956224  \n",
       "yay                  0.936053           0.946321  \n",
       "haha                 0.999505           0.939538  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from scipy.stats import hmean\n",
    "from scipy.stats import norm\n",
    "def normcdf(x):\n",
    "    return norm.cdf(x, x.mean(), x.std())\n",
    "\n",
    "neg = np.squeeze(np.asarray(neg_tf))\n",
    "pos = np.squeeze(np.asarray(pos_tf))\n",
    "term_freq_df2 = pd.DataFrame([neg,pos],columns=cvec.get_feature_names()).transpose()\n",
    "term_freq_df2.columns = ['negative', 'positive']\n",
    "term_freq_df2['total'] = term_freq_df2['negative'] + term_freq_df2['positive']\n",
    "term_freq_df2['pos_rate'] = term_freq_df2['positive'] * 1./term_freq_df2['total']\n",
    "term_freq_df2['pos_freq_pct'] = term_freq_df2['positive'] * 1./term_freq_df2['positive'].sum()\n",
    "term_freq_df2['pos_rate_normcdf'] = normcdf(term_freq_df2['pos_rate'])\n",
    "term_freq_df2['pos_freq_pct_normcdf'] = normcdf(term_freq_df2['pos_freq_pct'])\n",
    "term_freq_df2['pos_normcdf_hmean'] = hmean([term_freq_df2['pos_rate_normcdf'], term_freq_df2['pos_freq_pct_normcdf']])\n",
    "term_freq_df2.sort_values(by='pos_normcdf_hmean', ascending=False).iloc[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The calculation of the positivity score I decided is fairly simple and straightforward. For each word in a document, look it up in the list of 10,000 words I built vocabulary with, and get the corresponding 'pos_normcdf_hmean' value, then for the document calculate the average 'pos_normcdf_hmean' value. If none of the words can be found from the built 10,000 terms, then yields random probability ranging between 0 to 1. And the single value I get for a document is handled as a probability of the document being positive class."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Normally, a lexical approach will take many other aspects into the calculation to refine the prediction result, but I will try a very simple model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pos_hmean = term_freq_df2.pos_normcdf_hmean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.041663844300452516"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos_hmean['wtf']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_val_predicted_proba = []\n",
    "for t in x_validation:\n",
    "    hmean_scores = [pos_hmean[w] for w in t.split() if w in pos_hmean.index]\n",
    "    if len(hmean_scores) > 0:\n",
    "        prob_score = np.mean(hmean_scores)\n",
    "    else:\n",
    "        prob_score = np.random.random()\n",
    "    y_val_predicted_proba.append(prob_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pred = [1 if t > 0.56 else 0 for t in y_val_predicted_proba]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.75964912280701757"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "accuracy_score(y_validation,pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The accuracy is not as good as logistic regression with count vectorizer or  TFIDF vectorizer, but compared to null accuracy, 25.56% more accurate, and even compared to TextBlob sentiment analysis, my simple custom lexicon model is 15.31% more accurate. This is an impressive result for such a simple calculation and also considering the fact that the 'pos_normcdf_hmean' is calculated only with the training set. This might be useful later for an ensemble classifier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}