try_with_sampling.ipynb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#!/usr/bin/env python\n",
      "\n",
      "\"\"\"Porter Stemming Algorithm\n",
      "This is the Porter stemming algorithm, ported to Python from the\n",
      "version coded up in ANSI C by the author. It may be be regarded\n",
      "as canonical, in that it follows the algorithm presented in\n",
      "\n",
      "Porter, 1980, An algorithm for suffix stripping, Program, Vol. 14,\n",
      "no. 3, pp 130-137,\n",
      "\n",
      "only differing from it at the points maked --DEPARTURE-- below.\n",
      "\n",
      "See also http://www.tartarus.org/~martin/PorterStemmer\n",
      "\n",
      "The algorithm as described in the paper could be exactly replicated\n",
      "by adjusting the points of DEPARTURE, but this is barely necessary,\n",
      "because (a) the points of DEPARTURE are definitely improvements, and\n",
      "(b) no encoding of the Porter stemmer I have seen is anything like\n",
      "as exact as this version, even with the points of DEPARTURE!\n",
      "\n",
      "Vivake Gupta (v@nano.com)\n",
      "\n",
      "Release 1: January 2001\n",
      "\n",
      "Further adjustments by Santiago Bruno (bananabruno@gmail.com)\n",
      "to allow word input not restricted to one word per line, leading\n",
      "to:\n",
      "\n",
      "release 2: July 2008\n",
      "\"\"\"\n",
      "\n",
      "import sys\n",
      "\n",
      "class PorterStemmer:\n",
      "\n",
      "    def __init__(self):\n",
      "        \"\"\"The main part of the stemming algorithm starts here.\n",
      "        b is a buffer holding a word to be stemmed. The letters are in b[k0],\n",
      "        b[k0+1] ... ending at b[k]. In fact k0 = 0 in this demo program. k is\n",
      "        readjusted downwards as the stemming progresses. Zero termination is\n",
      "        not in fact used in the algorithm.\n",
      "\n",
      "        Note that only lower case sequences are stemmed. Forcing to lower case\n",
      "        should be done before stem(...) is called.\n",
      "        \"\"\"\n",
      "\n",
      "        self.b = \"\"  # buffer for word to be stemmed\n",
      "        self.k = 0\n",
      "        self.k0 = 0\n",
      "        self.j = 0   # j is a general offset into the string\n",
      "\n",
      "    def cons(self, i):\n",
      "        \"\"\"cons(i) is TRUE <=> b[i] is a consonant.\"\"\"\n",
      "        if self.b[i] == 'a' or self.b[i] == 'e' or self.b[i] == 'i' or self.b[i] == 'o' or self.b[i] == 'u':\n",
      "            return 0\n",
      "        if self.b[i] == 'y':\n",
      "            if i == self.k0:\n",
      "                return 1\n",
      "            else:\n",
      "                return (not self.cons(i - 1))\n",
      "        return 1\n",
      "\n",
      "    def m(self):\n",
      "        \"\"\"m() measures the number of consonant sequences between k0 and j.\n",
      "        if c is a consonant sequence and v a vowel sequence, and <..>\n",
      "        indicates arbitrary presence,\n",
      "\n",
      "           <c><v>       gives 0\n",
      "           <c>vc<v>     gives 1\n",
      "           <c>vcvc<v>   gives 2\n",
      "           <c>vcvcvc<v> gives 3\n",
      "           ....\n",
      "        \"\"\"\n",
      "        n = 0\n",
      "        i = self.k0\n",
      "        while 1:\n",
      "            if i > self.j:\n",
      "                return n\n",
      "            if not self.cons(i):\n",
      "                break\n",
      "            i = i + 1\n",
      "        i = i + 1\n",
      "        while 1:\n",
      "            while 1:\n",
      "                if i > self.j:\n",
      "                    return n\n",
      "                if self.cons(i):\n",
      "                    break\n",
      "                i = i + 1\n",
      "            i = i + 1\n",
      "            n = n + 1\n",
      "            while 1:\n",
      "                if i > self.j:\n",
      "                    return n\n",
      "                if not self.cons(i):\n",
      "                    break\n",
      "                i = i + 1\n",
      "            i = i + 1\n",
      "\n",
      "    def vowelinstem(self):\n",
      "        \"\"\"vowelinstem() is TRUE <=> k0,...j contains a vowel\"\"\"\n",
      "        for i in range(self.k0, self.j + 1):\n",
      "            if not self.cons(i):\n",
      "                return 1\n",
      "        return 0\n",
      "\n",
      "    def doublec(self, j):\n",
      "        \"\"\"doublec(j) is TRUE <=> j,(j-1) contain a double consonant.\"\"\"\n",
      "        if j < (self.k0 + 1):\n",
      "            return 0\n",
      "        if (self.b[j] != self.b[j-1]):\n",
      "            return 0\n",
      "        return self.cons(j)\n",
      "\n",
      "    def cvc(self, i):\n",
      "        \"\"\"cvc(i) is TRUE <=> i-2,i-1,i has the form consonant - vowel - consonant\n",
      "        and also if the second c is not w,x or y. this is used when trying to\n",
      "        restore an e at the end of a short  e.g.\n",
      "\n",
      "           cav(e), lov(e), hop(e), crim(e), but\n",
      "           snow, box, tray.\n",
      "        \"\"\"\n",
      "        if i < (self.k0 + 2) or not self.cons(i) or self.cons(i-1) or not self.cons(i-2):\n",
      "            return 0\n",
      "        ch = self.b[i]\n",
      "        if ch == 'w' or ch == 'x' or ch == 'y':\n",
      "            return 0\n",
      "        return 1\n",
      "\n",
      "    def ends(self, s):\n",
      "        \"\"\"ends(s) is TRUE <=> k0,...k ends with the string s.\"\"\"\n",
      "        length = len(s)\n",
      "        if s[length - 1] != self.b[self.k]: # tiny speed-up\n",
      "            return 0\n",
      "        if length > (self.k - self.k0 + 1):\n",
      "            return 0\n",
      "        if self.b[self.k-length+1:self.k+1] != s:\n",
      "            return 0\n",
      "        self.j = self.k - length\n",
      "        return 1\n",
      "\n",
      "    def setto(self, s):\n",
      "        \"\"\"setto(s) sets (j+1),...k to the characters in the string s, readjusting k.\"\"\"\n",
      "        length = len(s)\n",
      "        self.b = self.b[:self.j+1] + s + self.b[self.j+length+1:]\n",
      "        self.k = self.j + length\n",
      "\n",
      "    def r(self, s):\n",
      "        \"\"\"r(s) is used further down.\"\"\"\n",
      "        if self.m() > 0:\n",
      "            self.setto(s)\n",
      "\n",
      "    def step1ab(self):\n",
      "        \"\"\"step1ab() gets rid of plurals and -ed or -ing. e.g.\n",
      "\n",
      "           caresses  ->  caress\n",
      "           ponies    ->  poni\n",
      "           ties      ->  ti\n",
      "           caress    ->  caress\n",
      "           cats      ->  cat\n",
      "\n",
      "           feed      ->  feed\n",
      "           agreed    ->  agree\n",
      "           disabled  ->  disable\n",
      "\n",
      "           matting   ->  mat\n",
      "           mating    ->  mate\n",
      "           meeting   ->  meet\n",
      "           milling   ->  mill\n",
      "           messing   ->  mess\n",
      "\n",
      "           meetings  ->  meet\n",
      "        \"\"\"\n",
      "        if self.b[self.k] == 's':\n",
      "            if self.ends(\"sses\"):\n",
      "                self.k = self.k - 2\n",
      "            elif self.ends(\"ies\"):\n",
      "                self.setto(\"i\")\n",
      "            elif self.b[self.k - 1] != 's':\n",
      "                self.k = self.k - 1\n",
      "        if self.ends(\"eed\"):\n",
      "            if self.m() > 0:\n",
      "                self.k = self.k - 1\n",
      "        elif (self.ends(\"ed\") or self.ends(\"ing\")) and self.vowelinstem():\n",
      "            self.k = self.j\n",
      "            if self.ends(\"at\"):   self.setto(\"ate\")\n",
      "            elif self.ends(\"bl\"): self.setto(\"ble\")\n",
      "            elif self.ends(\"iz\"): self.setto(\"ize\")\n",
      "            elif self.doublec(self.k):\n",
      "                self.k = self.k - 1\n",
      "                ch = self.b[self.k]\n",
      "                if ch == 'l' or ch == 's' or ch == 'z':\n",
      "                    self.k = self.k + 1\n",
      "            elif (self.m() == 1 and self.cvc(self.k)):\n",
      "                self.setto(\"e\")\n",
      "\n",
      "    def step1c(self):\n",
      "        \"\"\"step1c() turns terminal y to i when there is another vowel in the stem.\"\"\"\n",
      "        if (self.ends(\"y\") and self.vowelinstem()):\n",
      "            self.b = self.b[:self.k] + 'i' + self.b[self.k+1:]\n",
      "\n",
      "    def step2(self):\n",
      "        \"\"\"step2() maps double suffices to single ones.\n",
      "        so -ization ( = -ize plus -ation) maps to -ize etc. note that the\n",
      "        string before the suffix must give m() > 0.\n",
      "        \"\"\"\n",
      "        if self.b[self.k - 1] == 'a':\n",
      "            if self.ends(\"ational\"):   self.r(\"ate\")\n",
      "            elif self.ends(\"tional\"):  self.r(\"tion\")\n",
      "        elif self.b[self.k - 1] == 'c':\n",
      "            if self.ends(\"enci\"):      self.r(\"ence\")\n",
      "            elif self.ends(\"anci\"):    self.r(\"ance\")\n",
      "        elif self.b[self.k - 1] == 'e':\n",
      "            if self.ends(\"izer\"):      self.r(\"ize\")\n",
      "        elif self.b[self.k - 1] == 'l':\n",
      "            if self.ends(\"bli\"):       self.r(\"ble\") # --DEPARTURE--\n",
      "            # To match the published algorithm, replace this phrase with\n",
      "            #   if self.ends(\"abli\"):      self.r(\"able\")\n",
      "            elif self.ends(\"alli\"):    self.r(\"al\")\n",
      "            elif self.ends(\"entli\"):   self.r(\"ent\")\n",
      "            elif self.ends(\"eli\"):     self.r(\"e\")\n",
      "            elif self.ends(\"ousli\"):   self.r(\"ous\")\n",
      "        elif self.b[self.k - 1] == 'o':\n",
      "            if self.ends(\"ization\"):   self.r(\"ize\")\n",
      "            elif self.ends(\"ation\"):   self.r(\"ate\")\n",
      "            elif self.ends(\"ator\"):    self.r(\"ate\")\n",
      "        elif self.b[self.k - 1] == 's':\n",
      "            if self.ends(\"alism\"):     self.r(\"al\")\n",
      "            elif self.ends(\"iveness\"): self.r(\"ive\")\n",
      "            elif self.ends(\"fulness\"): self.r(\"ful\")\n",
      "            elif self.ends(\"ousness\"): self.r(\"ous\")\n",
      "        elif self.b[self.k - 1] == 't':\n",
      "            if self.ends(\"aliti\"):     self.r(\"al\")\n",
      "            elif self.ends(\"iviti\"):   self.r(\"ive\")\n",
      "            elif self.ends(\"biliti\"):  self.r(\"ble\")\n",
      "        elif self.b[self.k - 1] == 'g': # --DEPARTURE--\n",
      "            if self.ends(\"logi\"):      self.r(\"log\")\n",
      "        # To match the published algorithm, delete this phrase\n",
      "\n",
      "    def step3(self):\n",
      "        \"\"\"step3() dels with -ic-, -full, -ness etc. similar strategy to step2.\"\"\"\n",
      "        if self.b[self.k] == 'e':\n",
      "            if self.ends(\"icate\"):     self.r(\"ic\")\n",
      "            elif self.ends(\"ative\"):   self.r(\"\")\n",
      "            elif self.ends(\"alize\"):   self.r(\"al\")\n",
      "        elif self.b[self.k] == 'i':\n",
      "            if self.ends(\"iciti\"):     self.r(\"ic\")\n",
      "        elif self.b[self.k] == 'l':\n",
      "            if self.ends(\"ical\"):      self.r(\"ic\")\n",
      "            elif self.ends(\"ful\"):     self.r(\"\")\n",
      "        elif self.b[self.k] == 's':\n",
      "            if self.ends(\"ness\"):      self.r(\"\")\n",
      "\n",
      "    def step4(self):\n",
      "        \"\"\"step4() takes off -ant, -ence etc., in context <c>vcvc<v>.\"\"\"\n",
      "        if self.b[self.k - 1] == 'a':\n",
      "            if self.ends(\"al\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'c':\n",
      "            if self.ends(\"ance\"): pass\n",
      "            elif self.ends(\"ence\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'e':\n",
      "            if self.ends(\"er\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'i':\n",
      "            if self.ends(\"ic\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'l':\n",
      "            if self.ends(\"able\"): pass\n",
      "            elif self.ends(\"ible\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'n':\n",
      "            if self.ends(\"ant\"): pass\n",
      "            elif self.ends(\"ement\"): pass\n",
      "            elif self.ends(\"ment\"): pass\n",
      "            elif self.ends(\"ent\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'o':\n",
      "            if self.ends(\"ion\") and (self.b[self.j] == 's' or self.b[self.j] == 't'): pass\n",
      "            elif self.ends(\"ou\"): pass\n",
      "            # takes care of -ous\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 's':\n",
      "            if self.ends(\"ism\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 't':\n",
      "            if self.ends(\"ate\"): pass\n",
      "            elif self.ends(\"iti\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'u':\n",
      "            if self.ends(\"ous\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'v':\n",
      "            if self.ends(\"ive\"): pass\n",
      "            else: return\n",
      "        elif self.b[self.k - 1] == 'z':\n",
      "            if self.ends(\"ize\"): pass\n",
      "            else: return\n",
      "        else:\n",
      "            return\n",
      "        if self.m() > 1:\n",
      "            self.k = self.j\n",
      "\n",
      "    def step5(self):\n",
      "        \"\"\"step5() removes a final -e if m() > 1, and changes -ll to -l if\n",
      "        m() > 1.\n",
      "        \"\"\"\n",
      "        self.j = self.k\n",
      "        if self.b[self.k] == 'e':\n",
      "            a = self.m()\n",
      "            if a > 1 or (a == 1 and not self.cvc(self.k-1)):\n",
      "                self.k = self.k - 1\n",
      "        if self.b[self.k] == 'l' and self.doublec(self.k) and self.m() > 1:\n",
      "            self.k = self.k -1\n",
      "\n",
      "    def stem(self, p, i, j):\n",
      "        \"\"\"In stem(p,i,j), p is a char pointer, and the string to be stemmed\n",
      "        is from p[i] to p[j] inclusive. Typically i is zero and j is the\n",
      "        offset to the last character of a string, (p[j+1] == '\\0'). The\n",
      "        stemmer adjusts the characters p[i] ... p[j] and returns the new\n",
      "        end-point of the string, k. Stemming never increases word length, so\n",
      "        i <= k <= j. To turn the stemmer into a module, declare 'stem' as\n",
      "        extern, and delete the remainder of this file.\n",
      "        \"\"\"\n",
      "        # copy the parameters into statics\n",
      "        self.b = p\n",
      "        self.k = j\n",
      "        self.k0 = i\n",
      "        if self.k <= self.k0 + 1:\n",
      "            return self.b # --DEPARTURE--\n",
      "\n",
      "        # With this line, strings of length 1 or 2 don't go through the\n",
      "        # stemming process, although no mention is made of this in the\n",
      "        # published algorithm. Remove the line to match the published\n",
      "        # algorithm.\n",
      "\n",
      "        self.step1ab()\n",
      "        self.step1c()\n",
      "        self.step2()\n",
      "        self.step3()\n",
      "        self.step4()\n",
      "        self.step5()\n",
      "        return self.b[self.k0:self.k+1]\n",
      "\n",
      "\n",
      "if __name__ == '__main__':\n",
      "    p = PorterStemmer()\n",
      "    word = 'raining'\n",
      "    print p.stem(word,0,len(word)-1)\n",
      "#    if len(sys.argv) > 1:\n",
      "#        for f in sys.argv[1:]:\n",
      "#            infile = open(f, 'r')\n",
      "#            while 1:\n",
      "#                output = ''\n",
      "#                word = ''\n",
      "#                line = infile.readline()\n",
      "#                if line == '':\n",
      "#                    break\n",
      "#                for c in line:\n",
      "#                    if c.isalpha():\n",
      "#                        word += c.lower()\n",
      "#                    else:\n",
      "#                        if word:\n",
      "#                            output += p.stem(word, 0,len(word)-1)\n",
      "#                            word = ''\n",
      "#                        output += c.lower()\n",
      "#                print output,\n",
      "#            infile.close()\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "rain\n"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#Author : Chitrasen\n",
      "#Date : 11/1/2013\n",
      "# Basic benchmark code LogisticRegression with sampling agents\n",
      "\n",
      "import pandas as pd\n",
      "import re\n",
      "from sklearn import cross_validation,metrics\n",
      "from sklearn.feature_extraction.text import CountVectorizer\n",
      "from sklearn import linear_model as lm\n",
      "from sklearn.metrics import mean_squared_error\n",
      "from sklearn.decomposition import TruncatedSVD\n",
      "from sklearn.tree import DecisionTreeClassifier\n",
      "from sklearn.multiclass import OneVsRestClassifier\n",
      "\n",
      "#import PorterStemmer\n",
      "\n",
      "def clean(s):\n",
      "    s = s.lower()\n",
      "    s = re.sub('n\\'t',' not',s)\n",
      "    s = re.sub('{link}',' ',s)\n",
      "    s = re.sub('#\\w+',' ',s)\n",
      "    s = re.sub('@\\w+',' ',s)\n",
      "    s = re.sub('[\\W]',' ',s)\n",
      "    s = re.sub('\\s+',' ',s)\n",
      "    s = removeStopWords(s)\n",
      "    return s\n",
      "\n",
      "def removeStopWords(sentance):\n",
      "    #stopWordsStr = 'a about above after again against all am an and any are as at be because been before being below between both but by cannot could did do does doing down during each few for from further had has have having he he\\'d he\\'ll he\\'s her here here\\'s hers herself him himself his how how\\'s i i\\'d i\\'ll i\\'m i\\'ve if in into is it it\\'s its itself let\\'s me more most my myself no nor not of off on once only or other ought our ours  ourselves out over own rt same she she\\'d she\\'ll she\\'s should so some such than that that\\'s the their theirs them themselves then there there\\'s these they they\\'d they\\'ll they\\'re they\\'ve this those through to too under until up very was we we\\'d we\\'ll we\\'re we\\'ve were what what\\'s when when\\'s where where\\'s which while who who\\'s whom why why\\'s with would you you\\'d you\\'ll you\\'re you\\'ve your yours yourself yourselves'\n",
      "    #stopWords = re.split('\\s+',stopWordsStr)\n",
      "    #cleanwords = [w for w in re.split('\\W+',s) if w not in stopWords]\n",
      "    \n",
      "    p = PorterStemmer()\n",
      "    \n",
      "    s = ''\n",
      "    for word in sentance.split('\\s+'):\n",
      "        s += \" \" +p.stem(word,0,len(word)-1)\n",
      "    return s\n",
      "\n",
      "\n",
      "def normalize_prob(a):\n",
      "    row_sums = a.sum(axis =1)\n",
      "    return a/row_sums[:,np.newaxis]\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "def isItRetweet(df):\n",
      "    return [len(re.findall('^rt', tweet)) > 0 for tweet in df.tweet]\n",
      "\n",
      "def hashTagsCount(df):\n",
      "    return [len(re.findall('\\#.',tweet)) for tweet in df.tweet]\n",
      "\n",
      "def mentionCount(df):\n",
      "    return [len(re.findall('\\@.',tweet)) for tweet in df.tweet]\n",
      "\n",
      "def numericContains(df):\n",
      "    return [len(re.findall('[0-9]+[:.]*[0-9]*.',tweet)) for tweet in df.tweet]\n",
      "    \n",
      "def containsAMPM(df):\n",
      "    return [len(re.findall('[0-9]+\\s*am|pm',tweet)) for tweet in df.tweet]\n",
      "    \n",
      "def linksCount(df):\n",
      "    return [len(re.findall('{link}',tweet)) for tweet in df.tweet]\n",
      "\n",
      "def containsHashWeather(df):\n",
      "    return [len(re.findall('\\#weather',tweet)) for tweet in df.tweet]\n",
      "\n",
      "def tweetLength(df):\n",
      "    return [len(tweet) for tweet in df.tweet]\n",
      "\n",
      "def getHighFeqWords(df,col,percent = 0.85, stop_words = 'english',max_features = 50):\n",
      "    df1 = df['tweet'][df[col] > percent]\n",
      "    countVect = CountVectorizer(strip_accents = 'ascii',analyzer='word',stop_words = stop_words,\n",
      "                                max_features = max_features,binary = True)    \n",
      "    \n",
      "    countVect.fit(df1.values)\n",
      "    maxFeatures = countVect.fit(df['tweet'][df[col] > percent]).get_feature_names()\n",
      "    print maxFeatures\n",
      "    countVect = CountVectorizer(vocabulary=maxFeatures)\n",
      "    return countVect.fit_transform(df.tweet).toarray()\n",
      "\n",
      "def getFeatureVecture(df,function_list = None,col_category=None):\n",
      "    \n",
      "    returnVals = None\n",
      "    \n",
      "    if(function_list == None):\n",
      "        function_chain = [isItRetweet,hashTagsCount,mentionCount,numericContains,containsAMPM,linksCount,\n",
      "                      containsHashWeather,tweetLength]\n",
      "        returnVals = np.hstack([np.matrix(func(df)).T for func in function_chain])\n",
      "        \n",
      "    return returnVals\n",
      "\n",
      "\n",
      "def main():\n",
      "    print 'Reading datasets...'\n",
      "    train = pd.read_csv('/home/trupti/data_local/kaggle/partly_sunny/train.csv',header = 0)\n",
      "    test = pd.read_csv('/home/trupti/data_local/kaggle/partly_sunny/test.csv', header = 0)\n",
      "\n",
      "\n",
      "    pred = pd.DataFrame(np.zeros((test.shape[0], train.shape[1]- test.shape[1])), columns = train.columns[4:])\n",
      "    pred.index = test['id']   \n",
      "\n",
      "\n",
      "    tweets_train = train['tweet']\n",
      "    tweets_test = test['tweet']\n",
      "    #X_all = list(tweets_train)+list(tweets_test)\n",
      "    \n",
      "    X_all_a = list(tweets_train)+list(tweets_test)\n",
      "    X_all = [clean(s) for s in X_all_a]\n",
      "    \n",
      "    tfidf = CountVectorizer(X_all,min_df=5,  max_features=None, strip_accents='unicode',  \n",
      "        analyzer='word',token_pattern=r'\\w{1,}',ngram_range=(1,1),stop_words = 'english')\n",
      "    \n",
      "    lr = lm.LogisticRegression(penalty='l2', dual=True, tol=0.0001, \n",
      "                           C=1, fit_intercept=True, intercept_scaling=1.0, \n",
      "                           class_weight=None, random_state=None)\n",
      "\n",
      "    clf = DecisionTreeClassifier()\n",
      "    \n",
      "    print 'fitting TfIDF...'\n",
      "    trainLen = tweets_train.shape[0]\n",
      "\n",
      "\n",
      "    tfidf.fit(X_all)\n",
      "    X_all = tfidf.transform(X_all)\n",
      "    lsa = TruncatedSVD(n_components = 100)\n",
      "\n",
      "    print 'Fitting in LSA'\n",
      "    X_all = lsa.fit_transform(X_all)\n",
      "\n",
      "    X_train = X_all[:trainLen]\n",
      "    X_test = X_all[trainLen:]\n",
      "    \n",
      "    X_train = np.hstack((X_train,getFeatureVecture(train)))\n",
      "\n",
      "\n",
      "    # sampling from the data and fitting indivisual samples linear model and average the results\n",
      "\n",
      "    final_cv_error = 0\n",
      "    sampling_count = 10\n",
      " \n",
      "    kf = cross_validation.KFold(X_train.shape[0],n_folds = 10)\n",
      "    cols = {'s':['s1','s2','s3','s4','s5'],\n",
      "        'w':['w1','w2','w3','w4'],\n",
      "        'k':['k1','k2','k3','k4','k5','k6','k7','k8','k9','k10','k11','k12','k13','k14','k15']}\n",
      "\n",
      "    col_keys = ['s','w','k']\n",
      "    final_error = []\n",
      "    j = 0\n",
      "\n",
      "    for train_idx,cv_idx in kf: \n",
      "        j +=1\n",
      "        print 'prcessing CV :',j\n",
      "        X_train_kf = X_train[train_idx,:]\n",
      "        X_cv_kf = X_train[cv_idx,:]\n",
      "        \n",
      "        y_train_kf = train.ix[train_idx,4:]\n",
      "        y_cv_kf = train.ix[cv_idx,4:]\n",
      "        #creating a place holder of cv predictions\n",
      "        pred_cv = pd.DataFrame(np.zeros(y_cv_kf.shape), columns = train.columns[4:])\n",
      "        \n",
      "        error_pred = 100\n",
      "        \n",
      "        for y_col in col_keys:\n",
      "            #do this for 100s simulation\n",
      "            \n",
      "            print '    Normalizing prediction values'\n",
      "            y_train_kf_norm = normalize_prob(y_train_kf[cols[y_col]].values)\n",
      "            \n",
      "            pred_sampling = np.zeros((len(cv_idx),sampling_count))\n",
      "            \n",
      "            #print 'sampling for ',y_col,sampling\n",
      "            sim_data = [np.random.choice(np.arange(len(cols[y_col])),\n",
      "                                     sampling_count,replace = True, \n",
      "                                     p = y_train_kf_norm[i,:]) for i in np.arange(len(train_idx))]\n",
      "            sim_data = np.array(sim_data)\n",
      "            print '    sim_data shape', sim_data.shape\n",
      "            #print sim_data[:10]\n",
      "            for sampling in range(sampling_count): \n",
      "                print '        processing sample :', sampling\n",
      "                onevsRest = OneVsRestClassifier(clf, n_jobs = -1)\n",
      "                onevsRest.fit(X_train_kf,sim_data[:,sampling])\n",
      "                pred_sampling[:,sampling] = onevsRest.predict(X_cv_kf)\n",
      "            pred_sampling = pd.DataFrame(pred_sampling)\n",
      "            \n",
      "            print '    processed all the samplings'\n",
      "            \n",
      "            col_index = 0\n",
      "            for col in cols[y_col]:            \n",
      "                col_val = pred_sampling == col_index            \n",
      "                pred_cv[col] = np.array(col_val.sum(axis = 1), dtype = float)/sampling_count\n",
      "                col_index +=1\n",
      "        print 'Score CV',j, sqrt(mean_squared_error(y_cv_kf,pred_cv))\n",
      "\n",
      "if __name__ == '__main__':\n",
      "    main()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Reading datasets...\n",
        "fitting TfIDF..."
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "Fitting in LSA"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "prcessing CV :"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 1\n",
        "    Normalizing prediction values"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    sim_data shape"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " (70151, 10)\n",
        "        processing sample : 0\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 1\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 2\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 3\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 4\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 5\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 6\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 7\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 8\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 9\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "    processed all the samplings\n",
        "    Normalizing prediction values\n",
        "    sim_data shape"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " (70151, 10)\n",
        "        processing sample : 0\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 1\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 2\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 3\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 4\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 5\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 6\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "        processing sample : 7\n"
       ]
      },
      {
       "ename": "KeyboardInterrupt",
       "evalue": "",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
        "\u001b[1;32m<ipython-input-13-de0622479e4b>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m    195\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    196\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0m__name__\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'__main__'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 197\u001b[1;33m     \u001b[0mmain\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
        "\u001b[1;32m<ipython-input-13-de0622479e4b>\u001b[0m in \u001b[0;36mmain\u001b[1;34m()\u001b[0m\n\u001b[0;32m    181\u001b[0m                 \u001b[1;32mprint\u001b[0m \u001b[1;34m'        processing sample :'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msampling\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    182\u001b[0m                 \u001b[0monevsRest\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mOneVsRestClassifier\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mn_jobs\u001b[0m \u001b[1;33m=\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;32m--> 183\u001b[1;33m                 \u001b[0monevsRest\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_train_kf\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0msim_data\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0msampling\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    184\u001b[0m                 \u001b[0mpred_sampling\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0msampling\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0monevsRest\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_cv_kf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    185\u001b[0m             \u001b[0mpred_sampling\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpred_sampling\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/usr/local/lib/python2.7/dist-packages/sklearn/multiclass.pyc\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, X, y)\u001b[0m\n\u001b[0;32m    199\u001b[0m         \"\"\"\n\u001b[0;32m    200\u001b[0m         self.estimators_, self.label_binarizer_ = fit_ovr(self.estimator, X, y,\n\u001b[1;32m--> 201\u001b[1;33m                                                           n_jobs=self.n_jobs)\n\u001b[0m\u001b[0;32m    202\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    203\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/usr/local/lib/python2.7/dist-packages/sklearn/multiclass.pyc\u001b[0m in \u001b[0;36mfit_ovr\u001b[1;34m(estimator, X, y, n_jobs)\u001b[0m\n\u001b[0;32m     90\u001b[0m     estimators = Parallel(n_jobs=n_jobs)(\n\u001b[0;32m     91\u001b[0m         \u001b[0mdelayed\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m_fit_binary\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mestimator\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[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mclasses\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m\"not %s\"\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 92\u001b[1;33m         for i in range(Y.shape[1]))\n\u001b[0m\u001b[0;32m     93\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mestimators\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlb\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     94\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/usr/local/lib/python2.7/dist-packages/sklearn/externals/joblib/parallel.pyc\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, iterable)\u001b[0m\n\u001b[0;32m    517\u001b[0m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdispatch\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfunction\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    518\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 519\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mretrieve\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    520\u001b[0m             \u001b[1;31m# Make sure that we get a last message telling us we are done\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    521\u001b[0m             \u001b[0melapsed_time\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[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_start_time\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/usr/local/lib/python2.7/dist-packages/sklearn/externals/joblib/parallel.pyc\u001b[0m in \u001b[0;36mretrieve\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    431\u001b[0m                             \u001b[1;31m# We can now allow subprocesses again\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    432\u001b[0m                             \u001b[0mos\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0menviron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'__JOBLIB_SPAWNED_PARALLEL__'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 433\u001b[1;33m                         \u001b[1;32mraise\u001b[0m \u001b[0mexception\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    434\u001b[0m                     \u001b[1;32melif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mexception\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mTransportableException\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    435\u001b[0m                         \u001b[1;31m# Capture exception to add information on the local\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
       ]
      }
     ],
     "prompt_number": 13
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#testing normalize_prob\n",
      "b = np.array([[.6,.4],[.1,.8],[.1,.4]])\n",
      "print b.shape\n",
      "print normalize_prob(b)\n",
      "np.sum(normalize_prob(b), axis = 1)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(3, 2)\n",
        "[[ 0.6         0.4       ]\n",
        " [ 0.11111111  0.88888889]\n",
        " [ 0.2         0.8       ]]\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 98,
       "text": [
        "array([ 1.,  1.,  1.])"
       ]
      }
     ],
     "prompt_number": 98
    }
   ],
   "metadata": {}
  }
 ]
}