imdbLSTM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSTM IMDB Movie Review Tutorial\n",
    "Josiah Olson"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "import time\n",
    "import numpy as np\n",
    "np.random.seed(1337)  # for reproducibility\n",
    "\n",
    "from keras.preprocessing import sequence\n",
    "from keras.models import Model, Sequential\n",
    "from keras.layers import Dense, Dropout, Embedding, LSTM, Input, merge, BatchNormalization\n",
    "from keras.datasets import imdb\n",
    "\n",
    "import os\n",
    "from keras.preprocessing.text import Tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "max_features = 10000\n",
    "max_len = 200  # cut texts after this number of words (among top max_features most common words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x:\n",
      "[\"Peter Ustinov plays an embezzler who is just getting out of prison when the film begins. As soon as he walks out the gates, he immediately begins working on a scheme to once again make a bundle by stealing, though this time he has his sights set pretty high. This is actually one of the weak points about the film, as he apparently knows nothing about computers (few did back in 1968) but manages to become a computer genius literally overnight! Yeah, right. Anyway, he comes up with a scheme to impersonate a computer expert and obtain a job with a large American corporation so he can eventually embezzle a ton of cash. Considering his knowledge of computers is rudimentary, it's amazing how he puts into effect a brilliant plan AND manages to infiltrate the computer system and its defenses. But, it's a movie after all, so I was able to suspend disbelief. By the end of the film, he and his new wife (Maggie Smith) are able to run away with a million pounds.<br /><br />At the very end, though, it gets very, very confusing and Smith announces she's managed to actually accumulate more than two million by shrewd investing in the companies that Ustinov started (though she didn't realize they were all dummy companies). This should mean that eventually these stocks she bought were worthless. What they seem to imply (and I could be guessing wrong here) is that Ustinov and his new partners quickly cashed in the stocks before this became known and the stocks would thereby then become worthless. Either way, the film seems to post on a magical ending whereby no one is hurt and everyone is happy--and this just didn't make much sense. It's a shame, really, as the acting and most of the writing was great. Karl Malden, Bob Newhart, Peter Ustinov and Maggie Smith were just wonderful.<br /><br />If I seem to have interpreted the end, let me know, as the film seemed very vague in details at the end.\"]\n",
      "['I understand this movie was made on a very low budget but that is no excuse for the monstrosity that is Grendel. Deathstalker, The Throne of Fire, Barbarian Queen, Conquest, the Invincible Barbarian were all done on shoestring budgets and poor special effects yet they still managed to create cult classics by adding some scantily clad women warriors and a good sense of humor. The primitive costumes, dark castles and beautiful Bulgarian landscape gave Grendel the potential to be a very good low budget sword and sorcery film, but the makers completely ruined this opportunity by using extremely poor CGI effects and colorless characters. Compare this film to Beowulf (1999). It may not be Citizen Kane but it is a good example of how an entertaining low budget sci-fi/ adventure movie can be made by using credible special effects and appealing characters.']\n",
      "25000\n",
      "y:\n",
      "[1]\n",
      "[0]\n",
      "25000\n"
     ]
    }
   ],
   "source": [
    "# get dataset and unzip: http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz\n",
    "\n",
    "X_train = []\n",
    "y_train = []\n",
    "\n",
    "path = './aclImdb/train/pos/'\n",
    "X_train.extend([open(path + f).read() for f in os.listdir(path) if f.endswith('.txt')])\n",
    "y_train.extend([1 for _ in range(12500)])\n",
    "\n",
    "path = './aclImdb/train/neg/'\n",
    "X_train.extend([open(path + f).read() for f in os.listdir(path) if f.endswith('.txt')])\n",
    "y_train.extend([0 for _ in range(12500)])\n",
    "\n",
    "print('x:')\n",
    "print(X_train[:1])\n",
    "print(X_train[-1:])\n",
    "print(len(X_train))\n",
    "print('y:')\n",
    "print(y_train[:1])\n",
    "print(y_train[-1:])\n",
    "print(len(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x:\n",
      "[\"This is most likely the best picture not many will see. It presented a culture in a real unhollywoodized way. A must see for all who like Indie and for those who don't. I think this movie will draw more into the Indie scene. The acting was top notch! The character of Alice was portreyed so well. With perfect akwardness. This movie ahould be brought to the mainstream! I think it would do phenominally! ALICE is the most real look at an element of our culture that I have seen since GO ASK ALICE. A look that is untouched by the Hollywood hand. Movies like this show young people that these things aren't glamorous but that they are real and compelling. If you liked PIECES OF APRIL you will love this one!\"]\n",
      "['During production, this LWT series was titled \\'Rocket To The Moon\\', a title that everyone on set at the time thought was cheesy enough. Then word came through that it had been retitled \\'Reach for the Moon\\', as if this the addition of this new word would evoke heavy metaphorical meanings involving the relationships and aspirations of the characters. This same heavy handed lack of subtlety and understanding is clear throughout the very fabric of the entire series, and the same tired and boring love-triangle storyline is actually stretched out over 7 episodes! Any television program that decides to use the \\'love-triangle\\' storyline for even a single episode is pushing it, but to smear it over seven episodes is unforgivable. There are reasons to watch however, with the scenery of the Isle of Wight certainly providing a beautiful setting, and the seemingly effortless performance of Lynda Bellingham hitting all the right comedy marks. This, however is not enough. A number of performances are noticeable , especially when placed against the static and emotionless wasteland which Jonathan Kerrigan refers to as his face. I have already mentioned Lynda Bellingham in the role of \\'fussy mother\\' (yes, thats as far as character development and back-story goes), but I was impressed by the actor Maurice Roeves (who is by far the most experienced actor in the cast, having appeared in everything from classic movies such as Richard Attenborough\\'s \\'Oh What a Lovely War\\', and \\'The Eagle Has Landed\\', and TV including Baywatch, Holby City, The Bill, Eastenders, Cheers, Doctor Who in the 80s, Star Trek, and even going to Hollywood with Sylvestor Stallone in 1995s Judge Dredd) whose performance reminded me very much of actual teachers I have met - often incapable men, whose lives are tinged with a very sad streak. It is refreshing to see that Roeves is still getting work, having also appeared in the 1998 British film \\'The Acid House\\' and even more recently alongside the wonderful Maria Bello and Sean Bean in \\'The Dark\\'. The standout reason to watch this series, even though his appearances are very few. Another actor who tries his hardest to elevate the program, is Ben Miles, in the role of the sex-mad \\'typical man\\' brother (once again, what an original character!). As with Roeves, the CV says it all. After elevating this, Miles has since appeared in the Hollywood movie \\'V for Vendetta\\', and the 2001 star studded film \\'The Affair of the Necklace\\', alongside Brian Cox, Adrien Brody, Christopher Walken and Jonathan Pryce. Unfortunately the fairly capable actors I have mentioned are resigned to relatively small roles in comparison to the dramatic driftwood that is Kerrigan and Saira Todd, both of whom seem to have been in Casualty for too long, and haven\\'t yet learned to act. Even the tolerable Frances Grey is made to suffer in the presence of such theatrically fetal mannequins, and unfortunately for her, the lasting memory I have of her from the set, is of the AD asking for another take of a relatively complex shot because during the take, she subconsciously \\'pulled her knickers out of her crack!\\'. Familiar faces appear in each episode - child actors from CBBC programs, and dramas such as \\'2 Point 4 Children\\', and other regulars from commercials such as Mr Ben\\'s (\"they\\'ll remember me for this\"), Lynda \\'the Oxo lady\\' Bellingham, and others that I don\\'t even want to remember. Each episode halfheartedly tries to \\'deal with issues\\' such as teenage pregnancy, disruptive pupils, and rocket building(!)... things that Grange Hill was doing better when my parents were young! Overall, this series was a disappointment, and a wasted opportunity to craft a genuinely interesting and well developed drama series. Unfortunately, for them they took the easy route, both in the writing and (for the most part) in the casting. I would recommend both \\'Waterloo Road\\' and \\'The Street\\',to fans of good drama - the latter in particular which managed to achieve this. If you like the Isle of Wight, don\\'t \\'reach for the moon\\', reach for the holiday brochure instead - it probably contains more character development!']\n",
      "25000\n",
      "y:\n",
      "[1]\n",
      "[0]\n",
      "25000\n"
     ]
    }
   ],
   "source": [
    "# read in the test data\n",
    "\n",
    "X_test = []\n",
    "y_test = []\n",
    "\n",
    "path = './aclImdb/test/pos/'\n",
    "X_test.extend([open(path + f).read() for f in os.listdir(path) if f.endswith('.txt')])\n",
    "y_test.extend([1 for _ in range(12500)])\n",
    "\n",
    "path = './aclImdb/test/neg/'\n",
    "X_test.extend([open(path + f).read() for f in os.listdir(path) if f.endswith('.txt')])\n",
    "y_test.extend([0 for _ in range(12500)])\n",
    "\n",
    "print('x:')\n",
    "print(X_test[:1])\n",
    "print(X_test[-1:])\n",
    "print(len(X_test))\n",
    "print('y:')\n",
    "print(y_test[:1])\n",
    "print(y_test[-1:])\n",
    "print(len(y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#tokenize works to list of integers where each integer is a key to a word\n",
    "imdbTokenizer = Tokenizer(nb_words=max_features)\n",
    "\n",
    "imdbTokenizer.fit_on_texts(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 and\n",
      "13 was\n",
      "18 but\n",
      "5 to\n",
      "9 it\n",
      "10 i\n",
      "17 movie\n",
      "4 of\n",
      "15 for\n",
      "6 is\n",
      "12 that\n",
      "7 br\n",
      "8 in\n",
      "19 film\n",
      "1 the\n",
      "14 as\n",
      "3 a\n",
      "11 this\n",
      "16 with\n"
     ]
    }
   ],
   "source": [
    "#print top 20 words \n",
    "#note zero is reserved for non frequent words\n",
    "for word, value in imdbTokenizer.word_index.items():\n",
    "    if value < 20:\n",
    "        print(value, word)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "the\n",
      "and\n",
      "an\n"
     ]
    }
   ],
   "source": [
    "#create int to word dictionary\n",
    "intToWord = {}\n",
    "for word, value in imdbTokenizer.word_index.items():\n",
    "    intToWord[value] = word\n",
    "\n",
    "#add a symbol for null placeholder\n",
    "intToWord[0] = \"!!!NA!!!\"\n",
    "    \n",
    "print(intToWord[1])\n",
    "print(intToWord[2])\n",
    "print(intToWord[32])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Peter Ustinov plays an embezzler who is just getting out of prison when the film begins. As soon as he walks out the gates, he immediately begins working on a scheme to once again make a bundle by stealing, though this time he has his sights set pretty high. This is actually one of the weak points about the film, as he apparently knows nothing about computers (few did back in 1968) but manages to become a computer genius literally overnight! Yeah, right. Anyway, he comes up with a scheme to impersonate a computer expert and obtain a job with a large American corporation so he can eventually embezzle a ton of cash. Considering his knowledge of computers is rudimentary, it's amazing how he puts into effect a brilliant plan AND manages to infiltrate the computer system and its defenses. But, it's a movie after all, so I was able to suspend disbelief. By the end of the film, he and his new wife (Maggie Smith) are able to run away with a million pounds.<br /><br />At the very end, though, it gets very, very confusing and Smith announces she's managed to actually accumulate more than two million by shrewd investing in the companies that Ustinov started (though she didn't realize they were all dummy companies). This should mean that eventually these stocks she bought were worthless. What they seem to imply (and I could be guessing wrong here) is that Ustinov and his new partners quickly cashed in the stocks before this became known and the stocks would thereby then become worthless. Either way, the film seems to post on a magical ending whereby no one is hurt and everyone is happy--and this just didn't make much sense. It's a shame, really, as the acting and most of the writing was great. Karl Malden, Bob Newhart, Peter Ustinov and Maggie Smith were just wonderful.<br /><br />If I seem to have interpreted the end, let me know, as the film seemed very vague in details at the end.\n",
      "[[823, 5304, 296, 32, 34, 6, 40, 394, 43, 4, 1168, 51, 1, 19, 775, 14, 512, 14, 26, 2367, 43, 1, 6322, 26, 1238, 775, 777, 20, 3, 4092, 5, 277, 171, 94, 3, 31, 3231, 148, 11, 55, 26, 44, 24, 7346, 267, 181, 309, 11, 6, 162, 28, 4, 1, 813, 754, 41, 1, 19, 14, 26, 681, 691, 161, 41, 5937, 168, 119, 142, 8, 4556, 18, 1025, 5, 410, 3, 1219, 1259, 1224, 9747, 1241, 205, 551, 26, 263, 53, 16, 3, 4092, 5, 3, 1219, 2800, 2, 5997, 3, 289, 16, 3, 1054, 295, 6468, 35, 26, 67, 847, 3, 5864, 4, 2207, 1066, 24, 1857, 4, 5937, 6, 42, 477, 86, 26, 1454, 80, 959, 3, 526, 1341, 2, 1025, 5, 1, 1219, 1507, 2, 91, 18, 42, 3, 17, 100, 29, 35, 10, 13, 499, 5, 4985, 2781, 31, 1, 127, 4, 1, 19, 26, 2, 24, 159, 319, 4142, 1256, 23, 499, 5, 518, 242, 16, 3, 1428, 5679, 7, 7, 30, 1, 52, 127, 148, 9, 211, 52, 52, 1496, 2, 1256, 438, 1316, 5, 162, 50, 71, 104, 1428, 31, 8, 1, 5121, 12, 5304, 643, 148, 56, 158, 920, 33, 68, 29, 5121, 11, 141, 381, 12, 847, 131, 56, 1244, 68, 3557, 48, 33, 303, 5, 8966, 2, 10, 97, 27, 3104, 352, 130, 6, 12, 5304, 2, 24, 159, 5824, 942, 8, 1, 156, 11, 874, 570, 2, 1, 59, 7606, 92, 410, 3557, 342, 93, 1, 19, 183, 5, 1204, 20, 3, 2537, 274, 54, 28, 6, 1486, 2, 313, 6, 651, 2, 11, 40, 158, 94, 73, 278, 42, 3, 899, 63, 14, 1, 113, 2, 88, 4, 1, 484, 13, 84, 5278, 8385, 2051, 823, 5304, 2, 4142, 1256, 68, 40, 386, 7, 7, 45, 10, 303, 5, 25, 9852, 1, 127, 384, 69, 121, 14, 1, 19, 465, 52, 3479, 8, 1372, 30, 1, 127]]\n",
      "peter\n",
      "ustinov\n",
      "plays\n",
      "an\n",
      "who\n",
      "is\n",
      "just\n",
      "getting\n",
      "out\n",
      "of\n",
      "prison\n",
      "when\n",
      "the\n",
      "film\n",
      "begins\n",
      "as\n",
      "soon\n",
      "as\n",
      "he\n",
      "walks\n",
      "out\n",
      "the\n",
      "gates\n",
      "he\n",
      "immediately\n",
      "begins\n",
      "working\n",
      "on\n",
      "a\n",
      "scheme\n",
      "to\n",
      "once\n",
      "again\n",
      "make\n",
      "a\n",
      "by\n",
      "stealing\n",
      "though\n",
      "this\n",
      "time\n",
      "he\n",
      "has\n",
      "his\n",
      "sights\n",
      "set\n",
      "pretty\n",
      "high\n",
      "this\n",
      "is\n",
      "actually\n",
      "one\n",
      "of\n",
      "the\n",
      "weak\n",
      "points\n",
      "about\n",
      "the\n",
      "film\n",
      "as\n",
      "he\n",
      "apparently\n",
      "knows\n",
      "nothing\n",
      "about\n",
      "computers\n",
      "few\n",
      "did\n",
      "back\n",
      "in\n",
      "1968\n",
      "but\n",
      "manages\n",
      "to\n",
      "become\n",
      "a\n",
      "computer\n",
      "genius\n",
      "literally\n",
      "overnight\n",
      "yeah\n",
      "right\n",
      "anyway\n",
      "he\n",
      "comes\n",
      "up\n",
      "with\n",
      "a\n",
      "scheme\n",
      "to\n",
      "a\n",
      "computer\n",
      "expert\n",
      "and\n",
      "obtain\n",
      "a\n",
      "job\n",
      "with\n",
      "a\n",
      "large\n",
      "american\n",
      "corporation\n",
      "so\n",
      "he\n",
      "can\n",
      "eventually\n",
      "a\n",
      "ton\n",
      "of\n",
      "cash\n",
      "considering\n",
      "his\n",
      "knowledge\n",
      "of\n",
      "computers\n",
      "is\n",
      "it's\n",
      "amazing\n",
      "how\n",
      "he\n",
      "puts\n",
      "into\n",
      "effect\n",
      "a\n",
      "brilliant\n",
      "plan\n",
      "and\n",
      "manages\n",
      "to\n",
      "the\n",
      "computer\n",
      "system\n",
      "and\n",
      "its\n",
      "but\n",
      "it's\n",
      "a\n",
      "movie\n",
      "after\n",
      "all\n",
      "so\n",
      "i\n",
      "was\n",
      "able\n",
      "to\n",
      "suspend\n",
      "disbelief\n",
      "by\n",
      "the\n",
      "end\n",
      "of\n",
      "the\n",
      "film\n",
      "he\n",
      "and\n",
      "his\n",
      "new\n",
      "wife\n",
      "maggie\n",
      "smith\n",
      "are\n",
      "able\n",
      "to\n",
      "run\n",
      "away\n",
      "with\n",
      "a\n",
      "million\n",
      "pounds\n",
      "br\n",
      "br\n",
      "at\n",
      "the\n",
      "very\n",
      "end\n",
      "though\n",
      "it\n",
      "gets\n",
      "very\n",
      "very\n",
      "confusing\n",
      "and\n",
      "smith\n",
      "she's\n",
      "managed\n",
      "to\n",
      "actually\n",
      "more\n",
      "than\n",
      "two\n",
      "million\n",
      "by\n",
      "in\n",
      "the\n",
      "companies\n",
      "that\n",
      "ustinov\n",
      "started\n",
      "though\n",
      "she\n",
      "didn't\n",
      "realize\n",
      "they\n",
      "were\n",
      "all\n",
      "companies\n",
      "this\n",
      "should\n",
      "mean\n",
      "that\n",
      "eventually\n",
      "these\n",
      "she\n",
      "bought\n",
      "were\n",
      "worthless\n",
      "what\n",
      "they\n",
      "seem\n",
      "to\n",
      "imply\n",
      "and\n",
      "i\n",
      "could\n",
      "be\n",
      "guessing\n",
      "wrong\n",
      "here\n",
      "is\n",
      "that\n",
      "ustinov\n",
      "and\n",
      "his\n",
      "new\n",
      "partners\n",
      "quickly\n",
      "in\n",
      "the\n",
      "before\n",
      "this\n",
      "became\n",
      "known\n",
      "and\n",
      "the\n",
      "would\n",
      "thereby\n",
      "then\n",
      "become\n",
      "worthless\n",
      "either\n",
      "way\n",
      "the\n",
      "film\n",
      "seems\n",
      "to\n",
      "post\n",
      "on\n",
      "a\n",
      "magical\n",
      "ending\n",
      "no\n",
      "one\n",
      "is\n",
      "hurt\n",
      "and\n",
      "everyone\n",
      "is\n",
      "happy\n",
      "and\n",
      "this\n",
      "just\n",
      "didn't\n",
      "make\n",
      "much\n",
      "sense\n",
      "it's\n",
      "a\n",
      "shame\n",
      "really\n",
      "as\n",
      "the\n",
      "acting\n",
      "and\n",
      "most\n",
      "of\n",
      "the\n",
      "writing\n",
      "was\n",
      "great\n",
      "karl\n",
      "malden\n",
      "bob\n",
      "peter\n",
      "ustinov\n",
      "and\n",
      "maggie\n",
      "smith\n",
      "were\n",
      "just\n",
      "wonderful\n",
      "br\n",
      "br\n",
      "if\n",
      "i\n",
      "seem\n",
      "to\n",
      "have\n",
      "interpreted\n",
      "the\n",
      "end\n",
      "let\n",
      "me\n",
      "know\n",
      "as\n",
      "the\n",
      "film\n",
      "seemed\n",
      "very\n",
      "vague\n",
      "in\n",
      "details\n",
      "at\n",
      "the\n",
      "end\n"
     ]
    }
   ],
   "source": [
    "#convert word strings to integer sequence lists\n",
    "print(X_train[0])\n",
    "print(imdbTokenizer.texts_to_sequences(X_train[:1]))\n",
    "for value in imdbTokenizer.texts_to_sequences(X_train[:1])[0]:\n",
    "    print(intToWord[value])\n",
    "    \n",
    "X_train = imdbTokenizer.texts_to_sequences(X_train)\n",
    "X_test = imdbTokenizer.texts_to_sequences(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25000 train sequences\n",
      "25000 test sequences\n",
      "Pad sequences (samples x time)\n",
      "X_train shape: (25000, 200)\n",
      "X_test shape: (25000, 200)\n"
     ]
    }
   ],
   "source": [
    "# Censor the data by having a max review length (in number of words)\n",
    "\n",
    "#use this function to load data from keras pickle instead of munging as shown above\n",
    "#(X_train, y_train), (X_test, y_test) = imdb.load_data(nb_words=max_features,\n",
    "#                                                      test_split=0.2)\n",
    "\n",
    "print(len(X_train), 'train sequences')\n",
    "print(len(X_test), 'test sequences')\n",
    "\n",
    "print(\"Pad sequences (samples x time)\")\n",
    "X_train = sequence.pad_sequences(X_train, maxlen=max_len)\n",
    "X_test = sequence.pad_sequences(X_test, maxlen=max_len)\n",
    "print('X_train shape:', X_train.shape)\n",
    "print('X_test shape:', X_test.shape)\n",
    "y_train = np.array(y_train)\n",
    "y_test = np.array(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x: [ 526 1341    2 1025    5    1 1219 1507    2   91   18   42    3   17  100\n",
      "   29   35   10   13  499    5 4985 2781   31    1  127    4    1   19   26\n",
      "    2   24  159  319 4142 1256   23  499    5  518  242   16    3 1428 5679\n",
      "    7    7   30    1   52  127  148    9  211   52   52 1496    2 1256  438\n",
      " 1316    5  162   50   71  104 1428   31    8    1 5121   12 5304  643  148\n",
      "   56  158  920   33   68   29 5121   11  141  381   12  847  131   56 1244\n",
      "   68 3557   48   33  303    5 8966    2   10   97   27 3104  352  130    6\n",
      "   12 5304    2   24  159 5824  942    8    1  156   11  874  570    2    1\n",
      "   59 7606   92  410 3557  342   93    1   19  183    5 1204   20    3 2537\n",
      "  274   54   28    6 1486    2  313    6  651    2   11   40  158   94   73\n",
      "  278   42    3  899   63   14    1  113    2   88    4    1  484   13   84\n",
      " 5278 8385 2051  823 5304    2 4142 1256   68   40  386    7    7   45   10\n",
      "  303    5   25 9852    1  127  384   69  121   14    1   19  465   52 3479\n",
      "    8 1372   30    1  127]\n",
      "y: 1\n"
     ]
    }
   ],
   "source": [
    "#example of a sentence sequence, note that lower integers are words that occur more commonly\n",
    "print(\"x:\", X_train[0]) #per observation vector of 20000 words\n",
    "print(\"y:\", y_train[0]) #positive or negative review encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y distribution: (array([0, 1]), array([12500, 12500]))\n",
      "max x word: 9999 ; min x word 0\n",
      "y distribution test: (array([0, 1]), array([12500, 12500]))\n",
      "max x word test: 9999 ; min x word 0\n"
     ]
    }
   ],
   "source": [
    "# double check that word sequences behave/final dimensions are as expected\n",
    "print(\"y distribution:\", np.unique(y_train, return_counts=True))\n",
    "print(\"max x word:\", np.max(X_train), \"; min x word\", np.min(X_train))\n",
    "print(\"y distribution test:\", np.unique(y_test, return_counts=True))\n",
    "print(\"max x word test:\", np.max(X_test), \"; min x word\", np.min(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "most and least popular words: \n",
      "(array([   0,    1,    2, ..., 9997, 9998, 9999], dtype=int32), array([1084248,  232308,  115674, ...,      23,      20,      22]))\n"
     ]
    }
   ],
   "source": [
    "print(\"most and least popular words: \")\n",
    "print(np.unique(X_train, return_counts=True))\n",
    "# as expected zero is the highly used word for words not in index"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What is an LSTM neural network?\n",
    "\n",
    "A Long Term Short Term Memory Network is a type of Recurrent Neural Network. RNN's have multiple time steps with a feature vecture input at each time step and the prior layer's output/hidden state as an input vector. RNN's come in many varieties, for example, an RNN can take one input and predict multiple outputs, multiple inputs and multiple outputs, multiple inputs to a single output dimension, ext.\n",
    "\n",
    "<img src=\"rnntypes.jpeg\">\n",
    "[source: Karpathy: The Unreasonable Effectiveness of Recurrent Neural Networks](http://karpathy.github.io/2015/05/21/rnn-effectiveness/)\n",
    "\n",
    "A more specific RNN where one-hot vector of English characters is the input and at character space in the text an input is provided and an output is generated for the predicted next character in the text.\n",
    "\n",
    "<img src=\"charseq.jpeg\">\n",
    "\n",
    "So, how is an RNN different from an LSTM? A normal RNN has the prior output concatenated with the current input to form the feature vector for the current layer. In a basic RNN structure long term dependencies (ie: the subject of a sentence) can be hard for the network to remember over many time steps. This is refered to as the vanishing/exploding gradient problem. More reading is suggested on this, but essentially the non-linearity's effect will compound over time step's causing the prior (historic) gradient to approach either zero or infinity.\n",
    "\n",
    "We provent this problem through the LSTM architecture. LSTM's have an input gate, output gate and forget gate. As simply as possible, this gates control what the output is for the current layer, what part of the past hidden state is forgotten based on the current input, and what part of the current input should be added to the networks long term memory vector which is passed from layer to layer along with the hidden state.\n",
    "\n",
    "<img src=\"lstmchain.png\">\n",
    "[source: Colah: Understanding LSTM Networks](http://colah.github.io/posts/2015-08-Understanding-LSTMs/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#set model hyper parameters\n",
    "epochs = 6\n",
    "embedding_neurons = 128\n",
    "lstm_neurons = 64\n",
    "batch_size = 32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                       Output Shape        Param #     Connected to                     \n",
      "====================================================================================================\n",
      "input_1 (InputLayer)               (None, 200)         0                                            \n",
      "____________________________________________________________________________________________________\n",
      "embedding_1 (Embedding)            (None, 200, 128)    1280000     input_1[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "batchnormalization_1 (BatchNormaliz(None, 200, 128)    256         embedding_1[0][0]                \n",
      "____________________________________________________________________________________________________\n",
      "lstm_1 (LSTM)                      (None, 64)          49408       batchnormalization_1[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)                (None, 64)          0           lstm_1[0][0]                     \n",
      "____________________________________________________________________________________________________\n",
      "dense_1 (Dense)                    (None, 1)           65          dropout_1[0][0]                  \n",
      "====================================================================================================\n",
      "Total params: 1329729\n",
      "____________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Forward Pass LSTM Network\n",
    "\n",
    "# this is the placeholder tensor for the input sequences\n",
    "sequence = Input(shape=(max_len,), dtype='int32')\n",
    "# this embedding layer will transform the sequences of integers\n",
    "# into vectors of size embedding\n",
    "# embedding layer converts dense int input to one-hot in real time to save memory\n",
    "embedded = Embedding(max_features, embedding_neurons, input_length=max_len)(sequence)\n",
    "# normalize embeddings by input/word in sentence\n",
    "bnorm = BatchNormalization()(embedded)\n",
    "\n",
    "# apply forwards LSTM layer size lstm_neurons\n",
    "forwards = LSTM(lstm_neurons, dropout_W=0.2, dropout_U=0.2)(bnorm)\n",
    "\n",
    "# dropout \n",
    "after_dp = Dropout(0.5)(forwards)\n",
    "output = Dense(1, activation='sigmoid')(after_dp)\n",
    "\n",
    "model_fdir_atom = Model(input=sequence, output=output)\n",
    "# review model structure\n",
    "print(model_fdir_atom.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train...\n",
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/6\n",
      "286s - loss: 0.5424 - acc: 0.7199 - val_loss: 0.3598 - val_acc: 0.8464\n",
      "Epoch 2/6\n",
      "287s - loss: 0.3132 - acc: 0.8713 - val_loss: 0.3283 - val_acc: 0.8644\n",
      "Epoch 3/6\n",
      "286s - loss: 0.2166 - acc: 0.9163 - val_loss: 0.3159 - val_acc: 0.8739\n",
      "Epoch 4/6\n",
      "286s - loss: 0.1548 - acc: 0.9433 - val_loss: 0.3610 - val_acc: 0.8717\n",
      "Epoch 5/6\n",
      "286s - loss: 0.1210 - acc: 0.9564 - val_loss: 0.3854 - val_acc: 0.8722\n",
      "Epoch 6/6\n",
      "288s - loss: 0.0950 - acc: 0.9660 - val_loss: 0.4609 - val_acc: 0.8709\n",
      "avg sec per epoch: 289.7011380990346\n"
     ]
    }
   ],
   "source": [
    "# Forward pass LSTM network\n",
    "\n",
    "# try using different optimizers and different optimizer configs\n",
    "model_fdir_atom.compile('adam', 'binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "print('Train...')\n",
    "start_time = time.time()\n",
    "\n",
    "history_fdir_atom = model_fdir_atom.fit(X_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    nb_epoch=epochs,\n",
    "                    validation_data=[X_test, y_test], \n",
    "                    verbose=2)\n",
    "\n",
    "end_time = time.time()\n",
    "average_time_per_epoch = (end_time - start_time) / epochs\n",
    "print(\"avg sec per epoch:\", average_time_per_epoch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A Bi-directional RNN is a network where the gradient is propagated both forward and backward through time. The hidden state and cell state vector is then the concatenation of the two time steps. This is useful because now the network knows about the input vectors (word embeddings) in front of and behind the current time step."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                       Output Shape        Param #     Connected to                     \n",
      "====================================================================================================\n",
      "input_2 (InputLayer)               (None, 200)         0                                            \n",
      "____________________________________________________________________________________________________\n",
      "embedding_2 (Embedding)            (None, 200, 128)    1280000     input_2[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "batchnormalization_2 (BatchNormaliz(None, 200, 128)    256         embedding_2[0][0]                \n",
      "____________________________________________________________________________________________________\n",
      "lstm_2 (LSTM)                      (None, 64)          49408       batchnormalization_2[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "lstm_3 (LSTM)                      (None, 64)          49408       batchnormalization_2[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "merge_1 (Merge)                    (None, 128)         0           lstm_2[0][0]                     \n",
      "                                                                   lstm_3[0][0]                     \n",
      "____________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)                (None, 128)         0           merge_1[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "dense_2 (Dense)                    (None, 1)           129         dropout_2[0][0]                  \n",
      "====================================================================================================\n",
      "Total params: 1379201\n",
      "____________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Bi-directional Atom\n",
    "\n",
    "# based on keras tutorial: https://github.com/fchollet/keras/blob/master/examples/imdb_bidirectional_lstm.py\n",
    "\n",
    "# this is the placeholder tensor for the input sequences\n",
    "sequence = Input(shape=(max_len,), dtype='int32')\n",
    "# this embedding layer will transform the sequences of integers\n",
    "# into vectors of size embedding\n",
    "# embedding layer converts dense int input to one-hot in real time to save memory\n",
    "embedded = Embedding(max_features, embedding_neurons, input_length=max_len)(sequence)\n",
    "# normalize embeddings by input/word in sentence\n",
    "bnorm = BatchNormalization()(embedded)\n",
    "\n",
    "# apply forwards LSTM layer size lstm_neurons\n",
    "forwards = LSTM(lstm_neurons, dropout_W=0.4, dropout_U=0.4)(bnorm)\n",
    "# apply backwards LSTM\n",
    "backwards = LSTM(lstm_neurons, dropout_W=0.4, dropout_U=0.4, go_backwards=True)(bnorm)\n",
    "\n",
    "# concatenate the outputs of the 2 LSTMs\n",
    "merged = merge([forwards, backwards], mode='concat', concat_axis=-1)\n",
    "after_dp = Dropout(0.5)(merged)\n",
    "output = Dense(1, activation='sigmoid')(after_dp)\n",
    "\n",
    "model_bidir_atom = Model(input=sequence, output=output)\n",
    "# review model structure\n",
    "print(model_bidir_atom.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train...\n",
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/6\n",
      "498s - loss: 0.5829 - acc: 0.6816 - val_loss: 0.4291 - val_acc: 0.8193\n",
      "Epoch 2/6\n",
      "506s - loss: 0.4177 - acc: 0.8143 - val_loss: 0.3763 - val_acc: 0.8534\n",
      "Epoch 3/6\n",
      "500s - loss: 0.3114 - acc: 0.8712 - val_loss: 0.3811 - val_acc: 0.8625\n",
      "Epoch 4/6\n",
      "504s - loss: 0.2598 - acc: 0.8978 - val_loss: 0.4055 - val_acc: 0.8445\n",
      "Epoch 5/6\n",
      "506s - loss: 0.2151 - acc: 0.9168 - val_loss: 0.3816 - val_acc: 0.8684\n",
      "Epoch 6/6\n",
      "507s - loss: 0.1810 - acc: 0.9305 - val_loss: 0.3958 - val_acc: 0.8664\n",
      "avg sec per epoch: 509.0115524133046\n"
     ]
    }
   ],
   "source": [
    "# Bi-directional Atom\n",
    "\n",
    "# try using different optimizers and different optimizer configs\n",
    "model_bidir_atom.compile('adam', 'binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "print('Train...')\n",
    "start_time = time.time()\n",
    "\n",
    "history_bidir_atom = model_bidir_atom.fit(X_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    nb_epoch=epochs,\n",
    "                    validation_data=[X_test, y_test], \n",
    "                    verbose=2)\n",
    "\n",
    "end_time = time.time()\n",
    "average_time_per_epoch = (end_time - start_time) / epochs\n",
    "print(\"avg sec per epoch:\", average_time_per_epoch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# run simple linear regression to compare performance\n",
    "\n",
    "#based on grid search done by: \n",
    "#https://github.com/rasbt/python-machine-learning-book/blob/master/code/ch08/ch08.ipynb\n",
    "\n",
    "#the tfidf vectors capture co-occurance statistics, think of each number representing how many times \n",
    "#a word occured in a text and scaled by word frequency\n",
    "\n",
    "tfidfTokenizer = Tokenizer(nb_words=max_features)\n",
    "tfidfTokenizer.fit_on_sequences(X_train.tolist())\n",
    "X_train_tfidf = np.asarray(tfidfTokenizer.sequences_to_matrix(X_train.tolist(), mode=\"tfidf\"))\n",
    "X_test_tfidf = np.asarray(tfidfTokenizer.sequences_to_matrix(X_test.tolist(), mode=\"tfidf\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.         -1.6183748  -1.73087555 ...,  0.          0.          0.        ]\n",
      " [ 0.         -1.71728432 -2.00060526 ...,  0.          0.          0.        ]\n",
      " [ 0.         -1.77371565 -2.33456537 ...,  0.          0.          0.        ]\n",
      " ..., \n",
      " [-0.62228544 -1.83609759 -2.20140072 ...,  0.          0.          0.        ]\n",
      " [-0.30308935 -2.31623382 -3.1618598  ...,  0.          0.          0.        ]\n",
      " [-0.78452538 -2.07616571 -2.09259727 ...,  0.          0.          0.        ]]\n",
      "(25000, 10000) (25000, 10000)\n"
     ]
    }
   ],
   "source": [
    "#check tfidf matrix\n",
    "print(X_train_tfidf)\n",
    "print(X_train_tfidf.shape, X_test_tfidf.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LibLinear]"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=0.001, 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=0, solver='liblinear', tol=0.0001,\n",
       "          verbose=1, warm_start=False)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "model_tfidf_reg = LogisticRegression(random_state=0, C=0.001, penalty='l2', verbose=1)\n",
    "model_tfidf_reg.fit(X_train_tfidf, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train acc: 0.9176\n",
      "test acc: 0.87328\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "#calculate test and train accuracy\n",
    "print(\"train acc:\", accuracy_score(y_test, model_tfidf_reg.predict(X_train_tfidf)))\n",
    "print(\"test acc:\", accuracy_score(y_test, model_tfidf_reg.predict(X_test_tfidf)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What went wrong? Why does our complex neural net not do leaps and bounds better than classic information retrieval techniques? Neural networks tend to excell with lot's of data. If we had around a million movie reviews the literature suggests that the LSTM would vastly outperform the logistic regression. \n",
    "\n",
    "<img src=\"rnnvridge.png\">\n",
    "[source: indico passage github](https://github.com/IndicoDataSolutions/Passage/tree/master/examples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                       Output Shape        Param #     Connected to                     \n",
      "====================================================================================================\n",
      "input_3 (InputLayer)               (None, 200)         0                                            \n",
      "____________________________________________________________________________________________________\n",
      "embedding_3 (Embedding)            (None, 200, 128)    1280000     input_3[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "batchnormalization_3 (BatchNormaliz(None, 200, 128)    256         embedding_3[0][0]                \n",
      "____________________________________________________________________________________________________\n",
      "lstm_4 (LSTM)                      (None, 64)          49408       batchnormalization_3[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "lstm_5 (LSTM)                      (None, 64)          49408       batchnormalization_3[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "merge_2 (Merge)                    (None, 128)         0           lstm_4[0][0]                     \n",
      "                                                                   lstm_5[0][0]                     \n",
      "____________________________________________________________________________________________________\n",
      "dropout_3 (Dropout)                (None, 128)         0           merge_2[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "dense_3 (Dense)                    (None, 1)           129         dropout_3[0][0]                  \n",
      "====================================================================================================\n",
      "Total params: 1379201\n",
      "____________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Bi-directional rmsprop\n",
    "\n",
    "# this example illistrate's that choice of optimizer is an important hyper-parameter for RNNs\n",
    "# rmsprop gives substancially better results than atom\n",
    "# in the literature these two optimizers commonly do well on RNNs\n",
    "\n",
    "# this is the placeholder tensor for the input sequences\n",
    "sequence = Input(shape=(max_len,), dtype='int32')\n",
    "# this embedding layer will transform the sequences of integers\n",
    "# into vectors of size embedding\n",
    "# embedding layer converts dense int input to one-hot in real time to save memory\n",
    "embedded = Embedding(max_features, embedding_neurons, input_length=max_len)(sequence)\n",
    "# normalize embeddings by input/word in sentence\n",
    "bnorm = BatchNormalization()(embedded)\n",
    "\n",
    "# apply forwards LSTM layer size lstm_neurons\n",
    "forwards = LSTM(lstm_neurons, dropout_W=0.4, dropout_U=0.4)(bnorm)\n",
    "# apply backwards LSTM\n",
    "backwards = LSTM(lstm_neurons, dropout_W=0.4, dropout_U=0.4, go_backwards=True)(bnorm)\n",
    "\n",
    "# concatenate the outputs of the 2 LSTMs\n",
    "merged = merge([forwards, backwards], mode='concat', concat_axis=-1)\n",
    "after_dp = Dropout(0.5)(merged)\n",
    "output = Dense(1, activation='sigmoid')(after_dp)\n",
    "\n",
    "model_bidir_rmsprop = Model(input=sequence, output=output)\n",
    "# review model structure\n",
    "print(model_bidir_rmsprop.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train...\n",
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/6\n",
      "523s - loss: 0.5914 - acc: 0.6716 - val_loss: 0.4526 - val_acc: 0.8062\n",
      "Epoch 2/6\n",
      "526s - loss: 0.3991 - acc: 0.8288 - val_loss: 0.3500 - val_acc: 0.8643\n",
      "Epoch 3/6\n",
      "526s - loss: 0.3101 - acc: 0.8750 - val_loss: 0.3185 - val_acc: 0.8796\n",
      "Epoch 4/6\n",
      "527s - loss: 0.2591 - acc: 0.8998 - val_loss: 0.3441 - val_acc: 0.8793\n",
      "Epoch 5/6\n",
      "517s - loss: 0.2206 - acc: 0.9159 - val_loss: 0.3362 - val_acc: 0.8776\n",
      "Epoch 6/6\n",
      "514s - loss: 0.1897 - acc: 0.9295 - val_loss: 0.3811 - val_acc: 0.8767\n",
      "avg sec per epoch: 527.796077132225\n"
     ]
    }
   ],
   "source": [
    "# Bi-directional rmsprop\n",
    "\n",
    "model_bidir_rmsprop.compile('rmsprop', 'binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "print('Train...')\n",
    "start_time = time.time()\n",
    "\n",
    "history_bidir_rmsprop = model_bidir_rmsprop.fit(X_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    nb_epoch=epochs,\n",
    "                    validation_data=[X_test, y_test], \n",
    "                    verbose=2)\n",
    "\n",
    "end_time = time.time()\n",
    "average_time_per_epoch = (end_time - start_time) / epochs\n",
    "print(\"avg sec per epoch:\", average_time_per_epoch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'trainable': True, 'input_length': 200, 'output_dim': 128, 'W_constraint': None, 'input_dtype': 'int32', 'init': 'uniform', 'mask_zero': False, 'W_regularizer': None, 'name': 'embedding_3', 'batch_input_shape': (None, 200), 'input_dim': 10000, 'dropout': 0.0, 'activity_regularizer': None}\n",
      "(10000, 128)\n"
     ]
    }
   ],
   "source": [
    "#get weights from embedding layer and visualize\n",
    "\n",
    "print(model_bidir_rmsprop.layers[1].get_config())\n",
    "embmatrix = model_bidir_rmsprop.layers[1].get_weights()[0]\n",
    "print(embmatrix.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5000, 2)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "topnwords = 5000\n",
    "toptsne = TSNE(n_components=2, random_state=0)\n",
    "tsneXY = toptsne.fit_transform(embmatrix[:topnwords, :]) \n",
    "tsneXY.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABZ0AAAWUCAYAAACqRqj7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAALoRnQEAAAAA6OY/AWjn4HG1oTFoAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa3881c4748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "displaytopnwords = 250\n",
    "import matplotlib.pyplot as plt\n",
    "fig, ax = plt.subplots()\n",
    "ax.scatter(tsneXY[:displaytopnwords, 0], tsneXY[:displaytopnwords, 1])\n",
    "\n",
    "for i in range(displaytopnwords):\n",
    "    ax.annotate(intToWord[i], (tsneXY[i, 0], tsneXY[i, 1]))\n",
    "\n",
    "fig.set_size_inches(25, 25)\n",
    "plt.show()\n",
    "# notice that great, most, well are clustered\n",
    "# bad don't even are clustered\n",
    "# We've learned structure in our sentiment embedding\n",
    "# neural networks give us this and other useful features for free"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note in the image above that 'script' is close to 'bad', apparently more people use that in the negative context. Also, 'performance' is close to 'him', it appears that the embedding is learning a reference back to the pronoun.\n",
    "Note, how far apart the clusters of positive and negative words are, negative words at the top of the TSNE embedding. Another cool observation, 'music' is close to 'feel', again think that aspect of this embedding is cinema specific."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(-7.2563256593202459, 3.9883630258994458, '!!!NA!!!')\n",
      "(1.7808394964577876, 0.28420900707493096, 'the')\n",
      "(1.3074923788709416, -1.413810383230826, 'and')\n",
      "(0.97258376513508604, -1.0636709051629747, 'a')\n",
      "(0.32762661520397507, -0.85173608614889351, 'of')\n",
      "(-1.4350479582962941, -0.80866758371634495, 'to')\n",
      "(3.4378501126279826, -0.18557639983027338, 'is')\n",
      "(-0.24889999686753927, -0.42062066515337743, 'br')\n",
      "(1.7262155260412964, -0.39519343021691189, 'in')\n",
      "(-3.6097427714182357, -2.0117301664713447, 'it')\n",
      "(-0.6329451082241131, -1.2518229318694318, 'i')\n",
      "(-1.7021053813659366, -0.73102130031323842, 'this')\n",
      "(-1.764117013490822, -0.7809036493439212, 'that')\n",
      "(-0.83022767719242463, 0.7586250481093747, 'was')\n",
      "(0.48896462079654518, -1.1895546496855136, 'as')\n",
      "(1.0988743098147409, 1.6877591127012967, 'for')\n",
      "(-3.1685738525530742, -0.62729307947154656, 'with')\n",
      "(-4.5218605670113332, -1.9000801332761281, 'movie')\n",
      "(4.7404464087564335, -1.7193498864520638, 'but')\n",
      "(4.8252599829559752, 0.96465357738898383, 'film')\n",
      "(-2.8345959772174814, -0.39800818426320017, 'on')\n",
      "(-3.0865337817632996, 4.2568253198803259, 'not')\n",
      "(0.47136785552726373, -2.5750364893749653, 'you')\n",
      "(1.1847985409385411, 0.24150841301422307, 'are')\n",
      "(3.272663355196932, -0.93469787200890775, 'his')\n",
      "(3.8694089689931928, 2.6220800331979568, 'have')\n",
      "(1.0073171039046291, 0.65736202629216334, 'he')\n",
      "(4.0787178502366173, 1.1497402031381456, 'be')\n",
      "(-1.8067958795468784, -0.74059367177157021, 'one')\n",
      "(-2.2053899022876049, 0.10199510256970566, 'all')\n",
      "(1.9016771590944137, 0.97163693440329346, 'at')\n",
      "(1.4369147281291941, 2.174057708998248, 'by')\n",
      "(2.0987808763272797, 1.0028642430926658, 'an')\n",
      "(-3.2792360803170686, -0.60599116000310849, 'they')\n",
      "(1.487272456141276, -1.0239365969149357, 'who')\n",
      "(1.6501811996470699, -0.084513267043402313, 'so')\n",
      "(2.9360007709128029, 2.7583635162824027, 'from')\n",
      "(-0.39362929220041165, -0.9023147303887723, 'like')\n",
      "(0.53308181941862498, -0.65258163071533226, 'her')\n",
      "(0.26942940478237914, -0.63916943456983522, 'or')\n",
      "(-5.4120715941874673, 5.1844360528411624, 'just')\n",
      "(0.67666118885775395, 2.3159268490938247, 'about')\n",
      "(0.6210945806814222, -2.9756192509272998, \"it's\")\n",
      "(0.22372442738831763, -0.54728912930576301, 'out')\n",
      "(3.6701692266959935, -0.14133169461724449, 'has')\n",
      "(4.4828206338447059, -0.42793962582541495, 'if')\n",
      "(3.8359168684525167, -0.53649396796861359, 'some')\n",
      "(-2.8545030889406777, -2.2767276658954207, 'there')\n",
      "(4.3561696020893992, 0.19919617635430484, 'what')\n",
      "(3.1356829563135578, -6.9534348348926107, 'good')\n",
      "(-2.1571806226082049, -2.5131431307465633, 'more')\n",
      "(-2.5567626990361165, -0.49597176162829404, 'when')\n",
      "(3.3388708971374759, -5.9229683739320027, 'very')\n",
      "(3.2136185005211244, -1.0137532755024645, 'up')\n",
      "(-1.9592448817129895, 6.5014183107809069, 'no')\n",
      "(5.0913974789202809, 1.5511679417868758, 'time')\n",
      "(3.9152515876020622, -4.1467148142536159, 'she')\n",
      "(0.66381075156722824, 5.8423022810968277, 'even')\n",
      "(2.0107236352081728, -1.0011659800296977, 'my')\n",
      "(-2.5650637236727452, 2.109802796172946, 'would')\n",
      "(4.105647818999226, 0.65397928533011362, 'which')\n",
      "(-2.0381243225483514, 4.5869878076481569, 'only')\n",
      "(-4.4635117688813342, -1.9467156229442, 'story')\n",
      "(-3.1083177572766441, -1.6891889967538714, 'really')\n",
      "(2.4534387500480954, -8.6228712647681647, 'see')\n",
      "(5.0463859841988246, 1.1248083473367099, 'their')\n",
      "(-5.0746864165396186, -0.75735325725481495, 'had')\n",
      "(1.1427499365559719, -0.090875588296553622, 'can')\n",
      "(-1.3173986472039589, 3.4805036147830606, 'were')\n",
      "(4.823406246557286, 0.95079435951443037, 'me')\n",
      "(-1.5297324388071885, -6.8987036454215174, 'well')\n",
      "(-1.952582442569412, -1.8988636351563637, 'than')\n",
      "(3.6185567291074285, 0.41628802699997397, 'we')\n",
      "(0.58042431663890992, 3.0518935939506382, 'much')\n",
      "(-0.46580129866972281, -3.7376029146911347, 'been')\n",
      "(-4.3045887922772748, 9.0547483844459844, 'bad')\n",
      "(3.1606696486019032, 0.035003845257118887, 'get')\n",
      "(3.4131259997682726, -0.75446859272851696, 'will')\n",
      "(4.199636456213387, 0.85005205984561316, 'do')\n",
      "(-1.8164992599579286, -6.5002572452739864, 'also')\n",
      "(1.8724742702044594, -0.009839759511626878, 'into')\n",
      "(3.8665064955289434, 1.8778096829918607, 'people')\n",
      "(0.61227868383135009, -4.0757374461608737, 'other')\n",
      "(2.4600753387357615, 0.039178917201315881, 'first')\n",
      "(-0.68938812575042163, -9.9352939590866232, 'great')\n",
      "(-0.48789226790178203, 1.4739449194876164, 'because')\n",
      "(3.5235873347931586, -2.0354139891315794, 'how')\n",
      "(-2.6728145687148959, -5.870750918059934, 'him')\n",
      "(1.2564554146337128, -5.5913953328884496, 'most')\n",
      "(-3.0846286130760299, 2.2621680632597383, \"don't\")\n",
      "(4.778982406537069, 1.2558123365187444, 'made')\n",
      "(1.8731228888614306, -0.79213868217823435, 'its')\n",
      "(0.58370408002254659, -0.51360376244934836, 'then')\n",
      "(4.5828788992030862, 1.1164116082206856, 'way')\n",
      "(-2.7741857805375365, -1.158393680201379, 'make')\n",
      "(4.5493567518322582, 1.3913183612457898, 'them')\n",
      "(-2.9194480488754051, -1.0279388629822217, 'too')\n",
      "(-1.3170475286472476, -0.55080219453032575, 'could')\n",
      "(-0.66825078126778592, 5.5977387547710169, 'any')\n",
      "(4.8215259763332066, 1.1632471306641972, 'movies')\n",
      "(1.4493086264317454, 0.18856560872949474, 'after')\n",
      "(3.4963970742967634, -2.6920903775165304, 'think')\n",
      "(3.9711250222375467, 2.347431061373908, 'characters')\n",
      "(5.2978777450222836, 1.0101882620654845, 'watch')\n",
      "(-4.3476123403124527, -0.58339877059601541, 'two')\n",
      "(-2.2668429788988051, -4.3194103547994924, 'films')\n",
      "(2.8909891532566094, 0.61273553771507017, 'character')\n",
      "(0.97429426091210136, -8.3222631181807927, 'seen')\n",
      "(3.9376247285741437, 0.61330121774578872, 'many')\n",
      "(-2.0779200287137845, 2.7482021846981484, 'being')\n",
      "(5.0756192846346337, -5.1263846799889174, 'life')\n",
      "(-2.476239211603279, 7.3100599776608615, 'plot')\n",
      "(-1.5964524829985023, -1.6216522182803665, 'never')\n",
      "(4.4470679289156649, 1.8443165201649883, 'acting')\n",
      "(-2.6756134871699468, -4.0112454088976204, 'little')\n",
      "(-0.77851455136049719, -9.7722851440229306, 'best')\n",
      "(-0.53883666776315509, -9.5751257425865486, 'love')\n",
      "(-2.224923971736922, 1.0078105204296885, 'over')\n",
      "(5.5355258119572799, -2.3402546835854667, 'where')\n",
      "(-0.68692276713259037, 3.3082317144423254, 'did')\n",
      "(1.3632010906695307, 1.9758520751751922, 'show')\n",
      "(-2.3730292702125082, -3.8462935395228248, 'know')\n",
      "(-3.3922302147002328, 1.7010179784532664, 'off')\n",
      "(4.8717457874688126, 1.8887969653201526, 'ever')\n",
      "(-1.06901816568443, -2.9156111652593282, 'does')\n",
      "(-2.4338548971362561, 7.3690897919569736, 'better')\n",
      "(0.45937800642844651, 1.2595446552324285, 'your')\n",
      "(0.35155459995930527, 2.8260170953402377, 'end')\n",
      "(3.7097382026170411, -5.5940195074977881, 'still')\n",
      "(-0.098243569566600672, -1.7827859208598735, 'man')\n",
      "(0.027513528127751097, 3.4902249121524926, 'here')\n",
      "(-0.29846568290875908, -0.25837200496049462, 'these')\n",
      "(3.7798663981035019, -0.94542871819033625, 'say')\n",
      "(4.6736193810298658, -4.3509820932621199, 'scene')\n",
      "(2.0040240762459947, 1.2629384719095016, 'while')\n",
      "(1.8165760854419462, 5.0151942277648951, 'why')\n",
      "(0.90589658440960719, 1.6574061612949746, 'scenes')\n",
      "(2.0800439172813436, -0.090730206113465217, 'go')\n",
      "(-0.10612525156035443, -1.4291976086651352, 'such')\n",
      "(-3.4794345767201982, 0.25797388990386605, 'something')\n",
      "(-5.3955359598189876, 4.4930874059815089, 'through')\n",
      "(2.4454277009806535, -0.074640490634138523, 'should')\n",
      "(4.5691460504784871, 0.93961312083966742, 'back')\n",
      "(3.3576854497857767, 3.5056929427206787, \"i'm\")\n",
      "(-3.1369331623106, -4.3787664845215382, 'real')\n",
      "(-2.8401133923799264, -0.92057079326778302, 'those')\n",
      "(1.403228075160021, 1.0940708438514546, 'watching')\n",
      "(-3.3364382002166231, -5.6311773086511883, 'now')\n",
      "(-2.2394941377807975, -3.1559670585392654, 'though')\n",
      "(-4.028887497070051, 6.1943501872046429, \"doesn't\")\n",
      "(-2.6297332278718306, -4.1535061318146704, 'years')\n",
      "(1.1711487812239421, -0.7321985306995088, 'thing')\n",
      "(-4.7757700282830147, -1.5512878586262135, 'old')\n",
      "(4.6534358883453262, 1.3887277983007165, 'actors')\n",
      "(3.3349319450117174, 2.1148539540983182, 'work')\n",
      "(5.3896431468920305, 1.0496812511647964, '10')\n",
      "(0.50336703389444448, 1.5344129103937028, 'before')\n",
      "(-3.7287583689379304, 1.3719479462060995, 'another')\n",
      "(-3.4449773027880086, 8.7973388132831349, \"didn't\")\n",
      "(1.7462142757853132, -1.6045438215853216, 'new')\n",
      "(-1.6408879543019579, -3.0984357704599925, 'funny')\n",
      "(-4.1281730210879495, 8.2330450744737664, 'nothing')\n",
      "(-3.2233682046800247, -2.0422547097548249, 'actually')\n",
      "(-0.32716246324185172, -4.0982107662349971, 'makes')\n",
      "(-3.444902111706627, 8.6609718512522402, 'director')\n",
      "(-3.5483509098555581, -1.706512293008644, 'look')\n",
      "(5.5175249920675471, -1.3774049916951649, 'find')\n",
      "(-0.91461609226171359, 4.5197697527000091, 'going')\n",
      "(-0.77616397355843192, 1.2417426629571986, 'few')\n",
      "(-3.9911369847007823, 1.6771467186240849, 'same')\n",
      "(2.3480555512728598, 0.67185117885980838, 'part')\n",
      "(5.1058275705683291, 1.5258438100319553, 'again')\n",
      "(4.0120765043872222, 1.2780225935371066, 'every')\n",
      "(-2.8373223048550109, -5.1269077481644967, 'lot')\n",
      "(3.1974794719694564, 2.018215233419411, 'cast')\n",
      "(-0.17162321412477224, -4.7418914248384718, 'us')\n",
      "(2.1161800297639695, -6.0729504665345537, 'quite')\n",
      "(0.55625632819386162, 2.3039072071715867, 'down')\n",
      "(-4.3441935166294092, -1.1520462262853146, 'want')\n",
      "(2.4387935496298363, -0.61797635672078766, 'world')\n",
      "(4.1575154776683849, -5.3147825336531209, 'things')\n",
      "(-2.9379723316462187, -0.57037592814116467, 'pretty')\n",
      "(-0.46389244434765003, -0.56665172451913792, 'young')\n",
      "(-2.9897806088517975, 3.8313200668780798, 'seems')\n",
      "(-2.659847962941162, -3.7764098414095266, 'around')\n",
      "(-0.99585135387503709, -0.40155216244246117, 'got')\n",
      "(-0.69285205050386678, 4.8414103488558791, 'horror')\n",
      "(-1.5131774793488673, -0.798592859899523, 'however')\n",
      "(1.9001520246985273, 1.8304225900247797, \"can't\")\n",
      "(-2.9768821110220101, -5.398029112656916, 'fact')\n",
      "(-3.5262318538435653, -5.1826650108990266, 'take')\n",
      "(4.2644195437615355, -0.29563691109165707, 'big')\n",
      "(-2.9713924643556191, 6.5834180837806731, 'enough')\n",
      "(0.070989665835133428, 3.3467348284988843, 'long')\n",
      "(2.4507947940830763, -4.7286897390471117, 'thought')\n",
      "(1.088081852193491, 1.2254540822397948, \"that's\")\n",
      "(-0.22531700444797131, -9.2415313823135925, 'both')\n",
      "(3.3150965785533049, -6.5738492712338559, 'between')\n",
      "(-1.8924478556711657, -4.5550417616008652, 'series')\n",
      "(0.78800434100930961, -0.51673664186421076, 'give')\n",
      "(2.1854666743553555, -1.7704754538151204, 'may')\n",
      "(-4.6092646541762514, 3.5816011348660881, 'original')\n",
      "(-3.5646834480190623, -4.5655317218150859, 'own')\n",
      "(3.4264852215092345, -4.7801695323617608, 'action')\n",
      "(4.7776740854131594, 1.66353622134561, \"i've\")\n",
      "(1.6736796182738261, -8.497286874344006, 'right')\n",
      "(2.4199881556494027, -0.20891773762335647, 'without')\n",
      "(1.5804632639188214, -8.0558034175579554, 'always')\n",
      "(2.0125000605724321, -8.3581686312890415, 'times')\n",
      "(5.0209679793674677, 0.38393208059763828, 'comedy')\n",
      "(-1.9812747601579748, 1.2883976266333754, 'point')\n",
      "(0.23929359509311834, 1.039662800957355, 'gets')\n",
      "(2.9736432572500755, -7.1345435794868797, 'must')\n",
      "(2.0568845047395898, -1.2209180003213649, 'come')\n",
      "(1.4579643317851447, 2.572532802689246, 'role')\n",
      "(-3.3405319896039223, 0.45326748454495852, \"isn't\")\n",
      "(4.6284728798607286, -4.3679097289692308, 'saw')\n",
      "(0.93977146490072061, -0.25789718657953203, 'almost')\n",
      "(-1.396041895812675, 6.5517448396980438, 'interesting')\n",
      "(-1.9196839214582997, 7.1374252776528389, 'least')\n",
      "(2.4360378242879719, -0.79324933001876774, 'family')\n",
      "(-3.0902107448484237, -4.5062565018629135, 'done')\n",
      "(-1.7872840326208137, 5.1224720982432386, \"there's\")\n",
      "(2.6787006803375064, 3.1825300008070565, 'whole')\n",
      "(1.1575079301795017, -5.8413015817337639, 'bit')\n",
      "(-4.2581461813347721, -2.2170891999041507, 'music')\n",
      "(-3.5937867350461765, 8.2881594675315373, 'script')\n",
      "(1.9115465582519742, 1.3405476486871741, 'far')\n",
      "(1.5258297653476849, 2.3051478865884776, 'making')\n",
      "(1.7966547277515719, 2.6540202386162868, 'guy')\n",
      "(-1.6876774583052769, 4.2598117797235338, 'anything')\n",
      "(-4.8032168075373223, -1.1549469330400621, 'feel')\n",
      "(-4.0659392064614401, 8.4159542799460887, 'minutes')\n",
      "(1.9453558312815658, 0.30438473346729278, 'last')\n",
      "(-2.6087061421960476, -4.8411477696578586, 'since')\n",
      "(-5.6663106842271604, 4.2774237453844348, 'might')\n",
      "(-3.2803020628473805, -6.1572393567162971, 'performance')\n",
      "(2.9246070631630761, -0.77091234295664179, \"he's\")\n",
      "(-3.0877835407451308, 5.6758860533606645, '2')\n",
      "(-0.78709889737184435, -0.95225324423877944, 'probably')\n",
      "(1.1991545262227974, 1.2361368552854377, 'kind')\n",
      "(-4.9196603604379945, 2.6116278396134391, 'am')\n",
      "(4.4479628118798766, 1.1922468287889121, 'away')\n",
      "(-3.7237129944101302, -4.034928642975097, 'yet')\n",
      "(-0.49407591821395119, 0.34385542513190365, 'rather')\n",
      "(0.57370173052576445, -0.49491699962118613, 'tv')\n",
      "(-4.5507212334304619, 9.3100125834068255, 'worst')\n",
      "(3.9572011881304778, -1.5780281225214701, 'girl')\n",
      "(-3.5489999220692123, -4.2106626940370369, 'day')\n",
      "(-3.3265189427866506, -4.7105857763074344, 'sure')\n"
     ]
    }
   ],
   "source": [
    "#guide to chart above\n",
    "\n",
    "for i in range(displaytopnwords):\n",
    "    print((tsneXY[i, 0], tsneXY[i, 1], intToWord[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('great', 0.0)\n",
      "('amazing', 0.5212827324867249)\n",
      "('wonderful', 0.5301917195320129)\n",
      "('impressed', 0.5487397909164429)\n",
      "('underrated', 0.5511659383773804)\n",
      "('beauty', 0.556002676486969)\n",
      "('awesome', 0.5607948899269104)\n",
      "('winner', 0.561858057975769)\n",
      "('holmes', 0.5652989745140076)\n",
      "('ride', 0.5676301121711731)\n",
      "('provoking', 0.5685350298881531)\n",
      "('sweet', 0.5689787864685059)\n",
      "('available', 0.568992018699646)\n",
      "('balance', 0.5692850351333618)\n",
      "('subtle', 0.5730404257774353)\n",
      "('criminal', 0.5735073685646057)\n",
      "('ironic', 0.5742459297180176)\n",
      "('unforgettable', 0.5747899413108826)\n",
      "('rare', 0.5760678052902222)\n",
      "('impact', 0.5766984224319458)\n"
     ]
    }
   ],
   "source": [
    "# Lets see what the embedding learned, \n",
    "# provoking is close to great in cosine space, that's cool and definettly movie specific\n",
    "\n",
    "from scipy.spatial.distance import euclidean\n",
    "\n",
    "for value in np.argsort(np.apply_along_axis(lambda x: euclidean(x, embmatrix[imdbTokenizer.word_index['great'],:]), \n",
    "                                            1, embmatrix))[:20]:\n",
    "    print((intToWord[value], euclidean(embmatrix[value,:], embmatrix[imdbTokenizer.word_index['great'],:])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('great', -7.119597689175805e-08)\n",
      "('amazing', 0.32049169159623303)\n",
      "('wonderful', 0.32156555138627929)\n",
      "('excellent', 0.37071652504353914)\n",
      "('available', 0.39210480941538683)\n",
      "('impressed', 0.39659269262071206)\n",
      "('today', 0.40157827429709725)\n",
      "('underrated', 0.40429114808571831)\n",
      "('ride', 0.40449692143994531)\n",
      "('beauty', 0.41162060546385293)\n",
      "('winner', 0.41534288040033673)\n",
      "('awesome', 0.41715241933314673)\n",
      "('heart', 0.41754174078939776)\n",
      "('perfect', 0.41879362639734119)\n",
      "('rare', 0.42205403276705222)\n",
      "('brilliant', 0.43154722189254757)\n",
      "('loved', 0.43247036209083212)\n",
      "('superb', 0.43305001243310526)\n",
      "('solid', 0.43443012027953165)\n",
      "('sweet', 0.43829286980697058)\n"
     ]
    }
   ],
   "source": [
    "from scipy.spatial.distance import cosine\n",
    "\n",
    "for value in np.argsort(np.apply_along_axis(lambda x: cosine(x, embmatrix[imdbTokenizer.word_index['great'],:]), \n",
    "                                            1, embmatrix))[:20]:\n",
    "    print((intToWord[value], cosine(embmatrix[value,:], embmatrix[imdbTokenizer.word_index['great'],:])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "84"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imdbTokenizer.word_index['great']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(25000, 200, 1)\n"
     ]
    }
   ],
   "source": [
    "# Prep the data to do a prediction at each time step\n",
    "# goal is to use this model to figure out what words cause predicted sentiment to change\n",
    "# reshape to predict at every time step the review sentiment\n",
    "\n",
    "y_train_multi = np.repeat(y_train.reshape((-1,1)), max_len, axis=1).reshape((-1,max_len,1))\n",
    "y_test_multi = np.repeat(y_test.reshape((-1,1)), max_len, axis=1).reshape((-1,max_len,1))\n",
    "print(y_train_multi.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                       Output Shape        Param #     Connected to                     \n",
      "====================================================================================================\n",
      "input_4 (InputLayer)               (None, 200)         0                                            \n",
      "____________________________________________________________________________________________________\n",
      "embedding_4 (Embedding)            (None, 200, 128)    1280000     input_4[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "batchnormalization_4 (BatchNormaliz(None, 200, 128)    256         embedding_4[0][0]                \n",
      "____________________________________________________________________________________________________\n",
      "lstm_6 (LSTM)                      (None, 200, 64)     49408       batchnormalization_4[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "dropout_4 (Dropout)                (None, 200, 64)     0           lstm_6[0][0]                     \n",
      "____________________________________________________________________________________________________\n",
      "timedistributed_1 (TimeDistributed)(None, 200, 1)      65          dropout_4[0][0]                  \n",
      "====================================================================================================\n",
      "Total params: 1329729\n",
      "____________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Forward Pass LSTM Network multi step predict\n",
    "from keras.layers.wrappers import TimeDistributed\n",
    "\n",
    "# this is the placeholder tensor for the input sequences\n",
    "sequence = Input(shape=(max_len,), dtype='int32')\n",
    "# this embedding layer will transform the sequences of integers\n",
    "# into vectors of size embedding\n",
    "# embedding layer converts dense int input to one-hot in real time to save memory\n",
    "embedded = Embedding(max_features, embedding_neurons, input_length=max_len)(sequence)\n",
    "# normalize embeddings by input/word in sentence\n",
    "bnorm = BatchNormalization()(embedded)\n",
    "\n",
    "# apply forwards LSTM layer size lstm_neurons\n",
    "forwards = LSTM(lstm_neurons, dropout_W=0.2, dropout_U=0.2, return_sequences=True)(bnorm)\n",
    "\n",
    "# dropout \n",
    "after_dp = Dropout(0.5)(forwards)\n",
    "output = TimeDistributed(Dense(1, activation='sigmoid'))(after_dp)\n",
    "\n",
    "model_fdir_multi = Model(input=sequence, output=output)\n",
    "# review model structure\n",
    "print(model_fdir_multi.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train...\n",
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/6\n",
      "328s - loss: 0.6050 - acc: 0.6518 - val_loss: 0.5505 - val_acc: 0.7073\n",
      "Epoch 2/6\n",
      "335s - loss: 0.4968 - acc: 0.7335 - val_loss: 0.5090 - val_acc: 0.7258\n",
      "Epoch 3/6\n",
      "343s - loss: 0.4503 - acc: 0.7593 - val_loss: 0.5221 - val_acc: 0.7252\n",
      "Epoch 4/6\n",
      "347s - loss: 0.4181 - acc: 0.7751 - val_loss: 0.5296 - val_acc: 0.7147\n",
      "Epoch 5/6\n",
      "347s - loss: 0.3945 - acc: 0.7865 - val_loss: 0.5916 - val_acc: 0.7173\n",
      "Epoch 6/6\n",
      "341s - loss: 0.3734 - acc: 0.7963 - val_loss: 0.5870 - val_acc: 0.7184\n",
      "avg sec per epoch: 343.41444504261017\n"
     ]
    }
   ],
   "source": [
    "# Forward pass LSTM network multi step predict\n",
    "\n",
    "# try using different optimizers and different optimizer configs\n",
    "model_fdir_multi.compile('rmsprop', 'binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "print('Train...')\n",
    "start_time = time.time()\n",
    "\n",
    "history_fdir_multi = model_fdir_multi.fit(X_train, y_train_multi,\n",
    "                    batch_size=batch_size,\n",
    "                    nb_epoch=epochs,\n",
    "                    validation_data=[X_test, y_test_multi], \n",
    "                    verbose=2)\n",
    "\n",
    "end_time = time.time()\n",
    "average_time_per_epoch = (end_time - start_time) / epochs\n",
    "print(\"avg sec per epoch:\", average_time_per_epoch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25000, 200, 1)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test_pred_mult = model_fdir_multi.predict(X_test)\n",
    "y_test_pred_mult.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.86119999999999997"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# as a sanity check look at the accuracy of the final prediction\n",
    "accuracy_score(y_test_multi[:,-1,:].ravel(), y_test_pred_mult[:,-1,:].ravel() > 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "avg starting review: 0.525825202465\n",
      "max starting review: 0.899692058563\n",
      "min starting review: 0.090793274343\n"
     ]
    }
   ],
   "source": [
    "print(\"avg starting review:\", np.median(y_test_pred_mult[:,0,:]))\n",
    "print(\"max starting review:\", np.max(y_test_pred_mult[:,0,:]))\n",
    "print(\"min starting review:\", np.min(y_test_pred_mult[:,0,:]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.50146971679277719"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(y_test_pred_mult[:,-1,:].ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.986877501011\n"
     ]
    }
   ],
   "source": [
    "for review in y_test_pred_mult[:,:,:][:10]:\n",
    "    for word in review:\n",
    "        print(word)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25000, 199, 1)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predDelta = y_test_pred_mult[:,1:,:] - y_test_pred_mult[:,:-1,:]\n",
    "predDelta.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25000, 200)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# concatenate 0.5 (neutral sentiment) to the initial dimension \n",
    "predDelta = np.concatenate((np.repeat(0.5, 25000).reshape((-1,1)), predDelta.reshape((-1, max_len-1))), axis=1)\n",
    "predDelta.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/thenome/anaconda3/lib/python3.5/site-packages/numpy/core/_methods.py:59: RuntimeWarning: Mean of empty slice.\n",
      "  warnings.warn(\"Mean of empty slice.\", RuntimeWarning)\n"
     ]
    }
   ],
   "source": [
    "# group the predDelta value by the sequence index value in X_test \n",
    "# to figure out which words cause sentiment to change the most\n",
    "\n",
    "ascDeltaWords = [np.mean(predDelta[X_test == x]) for x in range(max_features)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# filter out nan for words not observed in test set\n",
    "ascDeltaWords = [0 if np.isnan(x) else x for x in ascDeltaWords]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Larget positive deltas\n",
      "(318, 'excellent', 0.1135530759048801)\n",
      "(444, 'loved', 0.10824652494908857)\n",
      "(477, 'amazing', 0.1053537991351865)\n",
      "(774, 'fantastic', 0.096833608217064548)\n",
      "(526, 'brilliant', 0.091208578346822097)\n",
      "(767, 'surprised', 0.087269864032603492)\n",
      "(386, 'wonderful', 0.08310811262488664)\n",
      "(511, 'favorite', 0.0797732096012847)\n",
      "(817, 'realistic', 0.079194901447064689)\n",
      "(988, 'masterpiece', 0.075743083126526542)\n",
      "(507, 'enjoyed', 0.075512382078529103)\n",
      "(830, 'greatest', 0.073290431016200294)\n",
      "(400, 'perfect', 0.072729911018092799)\n",
      "(972, 'powerful', 0.071959564074604215)\n",
      "(639, 'hilarious', 0.069466818333931368)\n",
      "(115, 'best', 0.069015149195531311)\n",
      "(690, '7', 0.068242754817720061)\n",
      "(897, 'superb', 0.066684719748502846)\n",
      "(955, 'plenty', 0.06476325929448333)\n",
      "(734, 'enjoyable', 0.064571092524283752)\n",
      "(706, '8', 0.062462477409835196)\n"
     ]
    }
   ],
   "source": [
    "print(\"Larget positive deltas\")\n",
    "count = 0\n",
    "for value in np.argsort(ascDeltaWords).tolist()[::-1]:\n",
    "    #filter to only look at commonly used words\n",
    "    if value < 1000:\n",
    "        count += 1\n",
    "        print((value, intToWord[value], np.mean(predDelta[X_test == value])))\n",
    "        if count > 20:\n",
    "            break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Larget negative deltas\n",
      "(246, 'worst', -0.15909575486241459)\n",
      "(370, 'awful', -0.1244145495161994)\n",
      "(858, 'poorly', -0.10577680406002503)\n",
      "(469, 'unfortunately', -0.10169188567438291)\n",
      "(355, 'boring', -0.098990060330637994)\n",
      "(391, 'terrible', -0.097384153465498932)\n",
      "(832, 'lame', -0.0892755556459114)\n",
      "(335, 'poor', -0.082634137495456339)\n",
      "(524, 'horrible', -0.080569665198860843)\n",
      "(420, 'supposed', -0.080315845656580259)\n",
      "(613, 'annoying', -0.078367447825396883)\n",
      "(645, 'ridiculous', -0.077781594174762145)\n",
      "(750, 'dull', -0.076871606111156426)\n",
      "(944, 'mess', -0.074934110150582364)\n",
      "(75, 'bad', -0.0747984467143325)\n",
      "(907, 'badly', -0.074105306903817766)\n",
      "(860, 'okay', -0.071693332099578164)\n",
      "(375, 'stupid', -0.069879267600081313)\n",
      "(430, 'worse', -0.067640220894395683)\n",
      "(993, 'fails', -0.067170482558871147)\n",
      "(759, 'bunch', -0.065211902330915686)\n"
     ]
    }
   ],
   "source": [
    "print(\"Larget negative deltas\")\n",
    "count = 0\n",
    "for value in np.argsort(ascDeltaWords).tolist():\n",
    "    #filter to only look at commonly used words\n",
    "    if value < 1000:\n",
    "        count += 1\n",
    "        print((value, intToWord[value], np.mean(predDelta[X_test == value])))\n",
    "        if count > 20:\n",
    "            break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It's pretty cool that with only 25,000 training examples the network was able to learn so accurately what types of words cause a review to transition from positive to negative sentiment or vice versa. I have to image this would only improve with more data. \n",
    "\n",
    "Now let's concatenate the words with the highest absolute delta sentiment values and use the prior train TSNE embedding model to chart how these words are clustered on the manifold the network has learned."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABZ0AAAWUCAYAAACqRqj7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nABbzP59/aq3D68WHAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa22ff57978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "fig, ax = plt.subplots()\n",
    "ax.scatter(tsneXY[posToNegWordList, 0], tsneXY[posToNegWordList, 1])\n",
    "\n",
    "for i in posToNegWordList:\n",
    "    ax.annotate(intToWord[i], (tsneXY[i, 0], tsneXY[i, 1]))\n",
    "\n",
    "fig.set_size_inches(25, 25)\n",
    "plt.show()\n",
    "# notice that we have two very distinct clusters in the embedding as one would expect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next let's import pre-trained word vectors from google and use them to initialize our embedding to see if this improves the neural net's performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from gensim.models import Word2Vec\n",
    "\n",
    "#get pre trained word2vec from google:\n",
    "#https://doc-0k-4g-docs.googleusercontent.com/docs/securesc/gnqvgap6hjncpd3b10i2tv865io48jas/hmjtdgee48c14e1parufukrpkb8urra5/1463018400000/06848720943842814915/09676831593570546402/0B7XkCwpI5KDYNlNUTTlSS21pQmM?e=download&nonce=4l49745nmtine&user=09676831593570546402&hash=i2qe9mshan4mesl112ct9bu1tj9kr1hq\n",
    "\n",
    "googlew2v = Word2Vec.load_word2vec_format('./googleword2vec/GoogleNews-vectors-negative300.bin', binary=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.01353668  0.33976858  0.279984   ...,  0.14643192  0.16220523\n",
      "   0.34893078]\n",
      " [ 0.08007812  0.10498047  0.04980469 ...,  0.00366211  0.04760742\n",
      "  -0.06884766]\n",
      " [ 0.46788825  0.24908416  0.07994805 ...,  0.12019843  0.88012847\n",
      "   0.4712616 ]\n",
      " ..., \n",
      " [-0.28710938 -0.109375    0.17871094 ..., -0.09228516 -0.20605469\n",
      "   0.14550781]\n",
      " [-0.09082031  0.34960938  0.08154297 ...,  0.02502441 -0.24023438\n",
      "   0.14160156]\n",
      " [ 0.25976562  0.13574219  0.18847656 ..., -0.28515625  0.11474609\n",
      "   0.05688477]]\n",
      "(10000, 300)\n"
     ]
    }
   ],
   "source": [
    "# get word vectors for words in my index\n",
    "googleVecs = []\n",
    "for value in range(max_features):\n",
    "    try:\n",
    "        googleVecs.append(googlew2v[intToWord[value]])\n",
    "    except:\n",
    "        googleVecs.append(np.random.uniform(size=300))\n",
    "\n",
    "googleVecs = np.array(googleVecs)\n",
    "\n",
    "print(googleVecs)\n",
    "print(googleVecs.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                       Output Shape        Param #     Connected to                     \n",
      "====================================================================================================\n",
      "input_5 (InputLayer)               (None, 200)         0                                            \n",
      "____________________________________________________________________________________________________\n",
      "embedding_5 (Embedding)            (None, 200, 300)    3000000     input_5[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "batchnormalization_5 (BatchNormaliz(None, 200, 300)    600         embedding_5[0][0]                \n",
      "____________________________________________________________________________________________________\n",
      "lstm_7 (LSTM)                      (None, 64)          93440       batchnormalization_5[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "lstm_8 (LSTM)                      (None, 64)          93440       batchnormalization_5[0][0]       \n",
      "____________________________________________________________________________________________________\n",
      "merge_3 (Merge)                    (None, 128)         0           lstm_7[0][0]                     \n",
      "                                                                   lstm_8[0][0]                     \n",
      "____________________________________________________________________________________________________\n",
      "dropout_5 (Dropout)                (None, 128)         0           merge_3[0][0]                    \n",
      "____________________________________________________________________________________________________\n",
      "dense_5 (Dense)                    (None, 1)           129         dropout_5[0][0]                  \n",
      "====================================================================================================\n",
      "Total params: 3187609\n",
      "____________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Bi-directional google\n",
    "\n",
    "# this example tests if using pretrained embeddings will improve performance \n",
    "# relative to starting with random embeddings\n",
    "\n",
    "# this is the placeholder tensor for the input sequences\n",
    "sequence = Input(shape=(max_len,), dtype='int32')\n",
    "# this embedding layer will transform the sequences of integers\n",
    "# into vectors of size embedding\n",
    "# embedding layer converts dense int input to one-hot in real time to save memory\n",
    "embedded = Embedding(max_features, 300, input_length=max_len, weights=[googleVecs])(sequence)\n",
    "# normalize embeddings by input/word in sentence\n",
    "bnorm = BatchNormalization()(embedded)\n",
    "\n",
    "# apply forwards LSTM layer size lstm_neurons\n",
    "forwards = LSTM(lstm_neurons, dropout_W=0.4, dropout_U=0.4)(bnorm)\n",
    "# apply backwards LSTM\n",
    "backwards = LSTM(lstm_neurons, dropout_W=0.4, dropout_U=0.4, go_backwards=True)(bnorm)\n",
    "\n",
    "# concatenate the outputs of the 2 LSTMs\n",
    "merged = merge([forwards, backwards], mode='concat', concat_axis=-1)\n",
    "after_dp = Dropout(0.5)(merged)\n",
    "output = Dense(1, activation='sigmoid')(after_dp)\n",
    "\n",
    "model_bidir_google = Model(input=sequence, output=output)\n",
    "# review model structure\n",
    "print(model_bidir_google.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train...\n",
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/6\n",
      "718s - loss: 0.6614 - acc: 0.5989 - val_loss: 0.7493 - val_acc: 0.5954\n",
      "Epoch 2/6\n",
      "706s - loss: 0.5253 - acc: 0.7433 - val_loss: 0.4185 - val_acc: 0.8175\n",
      "Epoch 3/6\n",
      "685s - loss: 0.4361 - acc: 0.8084 - val_loss: 0.3547 - val_acc: 0.8581\n",
      "Epoch 4/6\n",
      "686s - loss: 0.3751 - acc: 0.8422 - val_loss: 0.3008 - val_acc: 0.8770\n",
      "Epoch 5/6\n",
      "678s - loss: 0.3290 - acc: 0.8668 - val_loss: 0.3027 - val_acc: 0.8825\n",
      "Epoch 6/6\n",
      "686s - loss: 0.3011 - acc: 0.8800 - val_loss: 0.2983 - val_acc: 0.8864\n",
      "avg sec per epoch: 699.137955625852\n"
     ]
    }
   ],
   "source": [
    "# Bi-directional google\n",
    "\n",
    "model_bidir_google.compile('rmsprop', 'binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "print('Train...')\n",
    "start_time = time.time()\n",
    "\n",
    "history_bidir_google = model_bidir_google.fit(X_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    nb_epoch=epochs,\n",
    "                    validation_data=[X_test, y_test], \n",
    "                    verbose=2)\n",
    "\n",
    "end_time = time.time()\n",
    "average_time_per_epoch = (end_time - start_time) / epochs\n",
    "print(\"avg sec per epoch:\", average_time_per_epoch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model with pre-trained word2vec vector's didn't outperform a random embedding, but I suspect it may have if I had run more epochs. Moreover, the number of parameters went up ~3x because the pre-trained vectors were 300d rather than 128d. I suspect I would have had better luck using the glove 6 billion vectors which are 100d instead. Our dataset is just too small to fit a 3 million weight embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.38526863]])"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test = \"I would want to see this movie.\"\n",
    "\n",
    "test = imdbTokenizer.texts_to_sequences([test])\n",
    "\n",
    "from keras.preprocessing import sequence\n",
    "test = sequence.pad_sequences(test, maxlen=max_len)\n",
    "\n",
    "test\n",
    "\n",
    "model_bidir_rmsprop.predict(test)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}