Emojifying Text.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python2.7/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import keras\n",
    "from keras.utils import to_categorical\n",
    "from keras.models import Model\n",
    "from keras.layers import Dense, Input, Dropout, LSTM, Activation\n",
    "from keras.layers.embeddings import Embedding\n",
    "from keras.preprocessing import sequence\n",
    "from keras.initializers import glorot_uniform\n",
    "from keras.models import Sequential"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Firstly , making word vector. Downloaded word vector online. Making a vector corresponding to each word from the word embediing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_glove_vector(path):\n",
    "    fh = open(path,\"r\")\n",
    "    lines=fh.readlines()\n",
    "    words=set()\n",
    "    word_to_vec={}\n",
    "    for element in lines:\n",
    "        modified_element=element.strip().split()\n",
    "        word=modified_element[0]\n",
    "        words.add(word)\n",
    "#         print word\n",
    "        word_to_vec[word]=np.array(modified_element[1:],dtype=np.float64)\n",
    "    \n",
    "    return words,word_to_vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "words,words_to_vec=read_glove_vector('glove.6B.50d.txt')#words is collection of unique words and word_to_vec is a dictionary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Word analogy using word emebeddings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check similarity of two vector using cosine similarity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The basic rule of finding analaogy between four words say a,b,c and d i.e. a is to b as c is to d is\n",
    "if we could find d such that embedding(a)-embedding(b)=embedding(c)-embedding(d)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cosine_similarity(u,v):\n",
    "    product=np.dot(u.T,v)\n",
    "    norm_u = np.sqrt(np.sum(np.square(u)))\n",
    "    norm_v = np.sqrt(np.sum(np.square(v)))\n",
    "    cosine_similarity = product/(norm_u*norm_v)\n",
    "    return cosine_similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Similarity between father and mother is 0.8909038442893615\n",
      "Similarity between father and india is 0.3332599608309279\n"
     ]
    }
   ],
   "source": [
    "print \"Similarity between father and mother is \"+ str(cosine_similarity(words_to_vec['father'],words_to_vec['mother']))\n",
    "print \"Similarity between father and india is \"+str(cosine_similarity(words_to_vec['father'],words_to_vec['india']))\n",
    "#notice below that Father and mother are more related than father and india"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def analogy(word_a,word_b,word_c):\n",
    "    word_a, word_b, word_c = word_a.lower(), word_b.lower(), word_c.lower()\n",
    "    input_vector=words_to_vec[word_a]-words_to_vec[word_b]#it is ea-eb basically\n",
    "    similarity=-100\n",
    "    word=\"\"\n",
    "    for i in words:\n",
    "        output_vector=words_to_vec[word_c]-words_to_vec[i]#it is ec-ed\n",
    "        score=cosine_similarity(input_vector,output_vector)\n",
    "        if score>similarity:\n",
    "            similarity=score\n",
    "            word=i\n",
    "    return word        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python2.7/dist-packages/ipykernel_launcher.py:5: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  \"\"\"\n"
     ]
    }
   ],
   "source": [
    "test_word=analogy('man','woman','boy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "As per analogy\n",
      "man is to woman as boy is  to girl\n"
     ]
    }
   ],
   "source": [
    "print \"As per analogy\"\n",
    "print \"man is to woman as boy is  to \"+test_word"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note we can use it for country capitals also"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python2.7/dist-packages/ipykernel_launcher.py:5: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  \"\"\"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "islamabad\n"
     ]
    }
   ],
   "source": [
    "print analogy(\"india\",'delhi','pakistan')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sentiment analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataframe=pd.read_csv('data/train_emoji.csv',names=['feature','label','col3','col4'])\n",
    "test_dataframe=pd.read_csv('data/tesss.csv',names=['feature','label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,Y_train=train_dataframe['feature'],train_dataframe['label']\n",
    "X_test,Y_test=test_dataframe['feature'],test_dataframe['label']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "food is life\n",
      "4\n"
     ]
    }
   ],
   "source": [
    "print X_train[4]\n",
    "print Y_train[4]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have some sentences and corresponding labels. We have 5 labels. Label=0,for emotions(affection),label=1, for sports related, label=2, for happiness, label=3, for sadness, label=4, for food."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y_oh_train=to_categorical(Y_train,num_classes=5).astype('float64')\n",
    "Y_oh_test=to_categorical(Y_test,num_classes=5).astype('float64')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reading glove vector "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "i = 1\n",
    "word_to_index = {}\n",
    "index_to_word = {}\n",
    "for w in sorted(words):\n",
    "    word_to_index[w] = i\n",
    "    index_to_word[i] = w\n",
    "    i = i + 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Using LSTM to do sentiment analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The basic idea is to use two layer LSTM model. Inputs will be the word vector. For example, if input is 'apple' we will use it's vector representation i.e. words_to_vec('apple') as input and so on.(Source: Andrew Ng's assignment)"
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![image.png](attachment:image.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The thing  with keras and infact with most deep learning frameworks is that for a given batch we want fixed length input. Say is one senetecne length is 4 and other is 6, we would need different time steps for the evaluation, which is somewhat not possible.Solution is to pad the input sentence to the length of maximum sentence"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "So, first step should be to convert sentence into it's indices. And then pad these sentences to lenth of longest sentence "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sentences_to_indices(X, word_to_index, max_len):\n",
    "    '''Converts array of sentences into array of indices'''\n",
    "    '''X is array of sentences of shape (m,1)'''\n",
    "    a=np.zeros((X.shape[0],max_len))\n",
    "    for i in range(X.shape[0]):\n",
    "        listofwords=X[i].lower().split()\n",
    "        for j in range(len(listofwords)):\n",
    "            a[i][j]=word_to_index[listofwords[j]]\n",
    "    return a     \n",
    "   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Making the Embedding() layer in Keras, using pre-trained word vectors. After this layer is built,  will pass the output of sentences_to_indices() to it as an input, and the Embedding() layer will return the word embeddings for a sentence. So, we can think of embedding layers as taking an input of shape(batchsize,maxlen) and outputting of shape(batchsize,maxlen,lengthofwordvector(in our case 50))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Making  an embedding layer in Keras"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It's like we are making a matrix in which corresponding to each index we have a word vector(of length 50). so, if we have 4lakh words than matrix should be of shape (4lakh,50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "maxLen = len(max(X_train, key=len).split())#maximum length of sentence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pretrained_embedding_layer(word_to_vec_map, word_to_index):\n",
    "    '''Will create an embedding layer and will load it with pretrained word embeddings'''\n",
    "    vocab_len=len(word_to_index)+1  #This will be row of embedded matrix\n",
    "    word_dim=word_to_vec_map['apple'].shape[0]\n",
    "    emb_matrix = np.zeros((vocab_len,word_dim))#Initializing embedded matrix\n",
    "    for word, index in word_to_index.items():\n",
    "        emb_matrix[index, :] = word_to_vec_map[word]\n",
    "    embedding_layer = Embedding(vocab_len, word_dim,input_length=maxLen)\n",
    "    \n",
    "    #Building the layer\n",
    "    embedding_layer.build((None,))\n",
    "    embedding_layer.set_weights([emb_matrix])\n",
    "    return embedding_layer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that our model will require input to be of shape (m,maxlen) and embedding layer will convert that to (m,maxlen,50). Now, see that embedding layer is outputting the same dimesion as expected by LSTM i.e. (m,tx,no of features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Defining model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "model=Sequential()\n",
    "embedding_layer = pretrained_embedding_layer(words_to_vec,word_to_index)\n",
    "model.add(embedding_layer)\n",
    "model.add(LSTM(128, return_sequences=True))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(LSTM(128))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(5,activation='softmax'))\n",
    "model.add(Activation('softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_1 (Embedding)      (None, 10, 50)            20000050  \n",
      "_________________________________________________________________\n",
      "lstm_1 (LSTM)                (None, 10, 128)           91648     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 10, 128)           0         \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, 128)               131584    \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 5)                 645       \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 5)                 0         \n",
      "=================================================================\n",
      "Total params: 20,223,927\n",
      "Trainable params: 20,223,927\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since our model expects \"X\" to be of shape (m,max_len).Convert 'x' to required shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_indices = sentences_to_indices(X_train, word_to_index, maxLen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.callbacks import ModelCheckpoint \n",
    "checkpointer = ModelCheckpoint(filepath='model.best.hdf5', \n",
    "                               verbose=1, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 118 samples, validate on 14 samples\n",
      "Epoch 1/50\n",
      "118/118 [==============================] - 4s 35ms/step - loss: 1.3860 - acc: 0.5339 - val_loss: 1.4640 - val_acc: 0.4286\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 1.46400, saving model to model.best.hdf5\n",
      "Epoch 2/50\n",
      "118/118 [==============================] - 4s 31ms/step - loss: 1.3446 - acc: 0.5678 - val_loss: 1.5249 - val_acc: 0.2857\n",
      "\n",
      "Epoch 00002: val_loss did not improve from 1.46400\n",
      "Epoch 3/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 1.2893 - acc: 0.6525 - val_loss: 1.3870 - val_acc: 0.5714\n",
      "\n",
      "Epoch 00003: val_loss improved from 1.46400 to 1.38697, saving model to model.best.hdf5\n",
      "Epoch 4/50\n",
      "118/118 [==============================] - 4s 30ms/step - loss: 1.2516 - acc: 0.6780 - val_loss: 1.4035 - val_acc: 0.5000\n",
      "\n",
      "Epoch 00004: val_loss did not improve from 1.38697\n",
      "Epoch 5/50\n",
      "118/118 [==============================] - 3s 30ms/step - loss: 1.2024 - acc: 0.7373 - val_loss: 1.3940 - val_acc: 0.5000\n",
      "\n",
      "Epoch 00005: val_loss did not improve from 1.38697\n",
      "Epoch 6/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 1.1595 - acc: 0.7542 - val_loss: 1.4307 - val_acc: 0.4286\n",
      "\n",
      "Epoch 00006: val_loss did not improve from 1.38697\n",
      "Epoch 7/50\n",
      "118/118 [==============================] - 5s 38ms/step - loss: 1.1313 - acc: 0.7797 - val_loss: 1.4002 - val_acc: 0.4286\n",
      "\n",
      "Epoch 00007: val_loss did not improve from 1.38697\n",
      "Epoch 8/50\n",
      "118/118 [==============================] - 3s 30ms/step - loss: 1.1275 - acc: 0.7797 - val_loss: 1.4231 - val_acc: 0.4286\n",
      "\n",
      "Epoch 00008: val_loss did not improve from 1.38697\n",
      "Epoch 9/50\n",
      "118/118 [==============================] - 4s 32ms/step - loss: 1.0985 - acc: 0.8136 - val_loss: 1.2695 - val_acc: 0.6429\n",
      "\n",
      "Epoch 00009: val_loss improved from 1.38697 to 1.26953, saving model to model.best.hdf5\n",
      "Epoch 10/50\n",
      "118/118 [==============================] - 3s 30ms/step - loss: 1.1082 - acc: 0.7881 - val_loss: 1.3048 - val_acc: 0.5714\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 1.26953\n",
      "Epoch 11/50\n",
      "118/118 [==============================] - 4s 31ms/step - loss: 1.0833 - acc: 0.8136 - val_loss: 1.2929 - val_acc: 0.6429\n",
      "\n",
      "Epoch 00011: val_loss did not improve from 1.26953\n",
      "Epoch 12/50\n",
      "118/118 [==============================] - 4s 32ms/step - loss: 1.0607 - acc: 0.8814 - val_loss: 1.1921 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00012: val_loss improved from 1.26953 to 1.19205, saving model to model.best.hdf5\n",
      "Epoch 13/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 1.0187 - acc: 0.9153 - val_loss: 1.2702 - val_acc: 0.6429\n",
      "\n",
      "Epoch 00013: val_loss did not improve from 1.19205\n",
      "Epoch 14/50\n",
      "118/118 [==============================] - 4s 32ms/step - loss: 1.0286 - acc: 0.8814 - val_loss: 1.2025 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 1.19205\n",
      "Epoch 15/50\n",
      "118/118 [==============================] - 3s 28ms/step - loss: 0.9733 - acc: 0.9576 - val_loss: 1.1292 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00015: val_loss improved from 1.19205 to 1.12922, saving model to model.best.hdf5\n",
      "Epoch 16/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 0.9715 - acc: 0.9576 - val_loss: 1.1365 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00016: val_loss did not improve from 1.12922\n",
      "Epoch 17/50\n",
      "118/118 [==============================] - 4s 30ms/step - loss: 0.9601 - acc: 0.9661 - val_loss: 1.1041 - val_acc: 0.8571\n",
      "\n",
      "Epoch 00017: val_loss improved from 1.12922 to 1.10408, saving model to model.best.hdf5\n",
      "Epoch 18/50\n",
      "118/118 [==============================] - 3s 28ms/step - loss: 0.9449 - acc: 0.9661 - val_loss: 1.0843 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00018: val_loss improved from 1.10408 to 1.08433, saving model to model.best.hdf5\n",
      "Epoch 19/50\n",
      "118/118 [==============================] - 5s 42ms/step - loss: 0.9510 - acc: 0.9576 - val_loss: 1.1821 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00019: val_loss did not improve from 1.08433\n",
      "Epoch 20/50\n",
      "118/118 [==============================] - 5s 39ms/step - loss: 0.9438 - acc: 0.9576 - val_loss: 1.1726 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00020: val_loss did not improve from 1.08433\n",
      "Epoch 21/50\n",
      "118/118 [==============================] - 4s 37ms/step - loss: 0.9257 - acc: 0.9831 - val_loss: 1.1514 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00021: val_loss did not improve from 1.08433\n",
      "Epoch 22/50\n",
      "118/118 [==============================] - 4s 32ms/step - loss: 0.9354 - acc: 0.9661 - val_loss: 1.0998 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00022: val_loss did not improve from 1.08433\n",
      "Epoch 23/50\n",
      "118/118 [==============================] - 4s 34ms/step - loss: 0.9333 - acc: 0.9746 - val_loss: 1.0801 - val_acc: 0.8571\n",
      "\n",
      "Epoch 00023: val_loss improved from 1.08433 to 1.08010, saving model to model.best.hdf5\n",
      "Epoch 24/50\n",
      "118/118 [==============================] - 4s 30ms/step - loss: 0.9331 - acc: 0.9746 - val_loss: 1.1650 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00024: val_loss did not improve from 1.08010\n",
      "Epoch 25/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 0.9334 - acc: 0.9831 - val_loss: 1.1640 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00025: val_loss did not improve from 1.08010\n",
      "Epoch 26/50\n",
      "118/118 [==============================] - 3s 30ms/step - loss: 0.9264 - acc: 0.9831 - val_loss: 1.1134 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00026: val_loss did not improve from 1.08010\n",
      "Epoch 27/50\n",
      "118/118 [==============================] - 4s 32ms/step - loss: 0.9359 - acc: 0.9746 - val_loss: 1.1180 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00027: val_loss did not improve from 1.08010\n",
      "Epoch 28/50\n",
      "118/118 [==============================] - 3s 30ms/step - loss: 0.9400 - acc: 0.9661 - val_loss: 1.1179 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00028: val_loss did not improve from 1.08010\n",
      "Epoch 29/50\n",
      "118/118 [==============================] - 4s 36ms/step - loss: 0.9411 - acc: 0.9661 - val_loss: 1.1885 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00029: val_loss did not improve from 1.08010\n",
      "Epoch 30/50\n",
      "118/118 [==============================] - 4s 33ms/step - loss: 0.9409 - acc: 0.9661 - val_loss: 1.3690 - val_acc: 0.5000\n",
      "\n",
      "Epoch 00030: val_loss did not improve from 1.08010\n",
      "Epoch 31/50\n",
      "118/118 [==============================] - 5s 42ms/step - loss: 0.9727 - acc: 0.9322 - val_loss: 1.0494 - val_acc: 0.8571\n",
      "\n",
      "Epoch 00031: val_loss improved from 1.08010 to 1.04938, saving model to model.best.hdf5\n",
      "Epoch 32/50\n",
      "118/118 [==============================] - 5s 40ms/step - loss: 0.9369 - acc: 0.9746 - val_loss: 1.0678 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00032: val_loss did not improve from 1.04938\n",
      "Epoch 33/50\n",
      "118/118 [==============================] - 4s 30ms/step - loss: 0.9584 - acc: 0.9492 - val_loss: 1.2610 - val_acc: 0.6429\n",
      "\n",
      "Epoch 00033: val_loss did not improve from 1.04938\n",
      "Epoch 34/50\n",
      "118/118 [==============================] - 5s 42ms/step - loss: 1.0746 - acc: 0.8305 - val_loss: 1.4681 - val_acc: 0.4286\n",
      "\n",
      "Epoch 00034: val_loss did not improve from 1.04938\n",
      "Epoch 35/50\n",
      "118/118 [==============================] - 4s 31ms/step - loss: 1.1151 - acc: 0.7797 - val_loss: 1.2607 - val_acc: 0.6429\n",
      "\n",
      "Epoch 00035: val_loss did not improve from 1.04938\n",
      "Epoch 36/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 1.0908 - acc: 0.8136 - val_loss: 1.1983 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00036: val_loss did not improve from 1.04938\n",
      "Epoch 37/50\n",
      "118/118 [==============================] - 3s 29ms/step - loss: 1.1138 - acc: 0.7881 - val_loss: 1.1855 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00037: val_loss did not improve from 1.04938\n",
      "Epoch 38/50\n",
      "118/118 [==============================] - 4s 37ms/step - loss: 1.0701 - acc: 0.8220 - val_loss: 1.3122 - val_acc: 0.5714\n",
      "\n",
      "Epoch 00038: val_loss did not improve from 1.04938\n",
      "Epoch 39/50\n",
      "118/118 [==============================] - 5s 40ms/step - loss: 1.0764 - acc: 0.8220 - val_loss: 1.3393 - val_acc: 0.5714\n",
      "\n",
      "Epoch 00039: val_loss did not improve from 1.04938\n",
      "Epoch 40/50\n",
      "118/118 [==============================] - 5s 46ms/step - loss: 1.0227 - acc: 0.8898 - val_loss: 1.1885 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00040: val_loss did not improve from 1.04938\n",
      "Epoch 41/50\n",
      "118/118 [==============================] - 5s 45ms/step - loss: 0.9656 - acc: 0.9407 - val_loss: 1.1275 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00041: val_loss did not improve from 1.04938\n",
      "Epoch 42/50\n",
      "118/118 [==============================] - 5s 45ms/step - loss: 0.9621 - acc: 0.9492 - val_loss: 1.1887 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00042: val_loss did not improve from 1.04938\n",
      "Epoch 43/50\n",
      "118/118 [==============================] - 4s 34ms/step - loss: 0.9486 - acc: 0.9661 - val_loss: 1.1741 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00043: val_loss did not improve from 1.04938\n",
      "Epoch 44/50\n",
      "118/118 [==============================] - 5s 43ms/step - loss: 0.9258 - acc: 0.9746 - val_loss: 1.2539 - val_acc: 0.6429\n",
      "\n",
      "Epoch 00044: val_loss did not improve from 1.04938\n",
      "Epoch 45/50\n",
      "118/118 [==============================] - 4s 31ms/step - loss: 0.9418 - acc: 0.9661 - val_loss: 1.1589 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00045: val_loss did not improve from 1.04938\n",
      "Epoch 46/50\n",
      "118/118 [==============================] - 4s 30ms/step - loss: 0.9397 - acc: 0.9661 - val_loss: 1.0627 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00046: val_loss did not improve from 1.04938\n",
      "Epoch 47/50\n",
      "118/118 [==============================] - 4s 37ms/step - loss: 0.9639 - acc: 0.9492 - val_loss: 1.0836 - val_acc: 0.7857\n",
      "\n",
      "Epoch 00047: val_loss did not improve from 1.04938\n",
      "Epoch 48/50\n",
      "118/118 [==============================] - 4s 36ms/step - loss: 0.9647 - acc: 0.9407 - val_loss: 1.1743 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00048: val_loss did not improve from 1.04938\n",
      "Epoch 49/50\n",
      "118/118 [==============================] - 4s 38ms/step - loss: 0.9443 - acc: 0.9576 - val_loss: 1.1889 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00049: val_loss did not improve from 1.04938\n",
      "Epoch 50/50\n",
      "118/118 [==============================] - 6s 49ms/step - loss: 0.9394 - acc: 0.9661 - val_loss: 1.1716 - val_acc: 0.7143\n",
      "\n",
      "Epoch 00050: val_loss did not improve from 1.04938\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fa7c7913e10>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train_indices, Y_oh_train, epochs = 50, batch_size = 32, shuffle=True,callbacks=[checkpointer],validation_split=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights('model.best.hdf5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "56/56 [==============================] - 1s 12ms/step\n",
      "('Test accuracy = ', 0.7678571343421936)\n"
     ]
    }
   ],
   "source": [
    "X_test_indices = sentences_to_indices(X_test, word_to_index, max_len = maxLen)\n",
    "loss, acc = model.evaluate(X_test_indices, Y_oh_test)\n",
    "print(\"Test accuracy = \", acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Emojifiying Text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "import emoji"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "dictionary={\"0\": ':heart:',\"1\": \":baseball:\",\"2\": \":smile:\",\"3\": \":disappointed:\",\"4\": \":fork_and_knife:\"}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "def emojify_text(label):\n",
    "    return emoji.emojize(dictionary[str(label)], use_aliases=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_sentence=np.array([\"I am not feeling good\",\"I want to go play some cricket\",'i want to go on a romantic date'])\n",
    "input_sentence_index=sentences_to_indices(input_sentence,word_to_index,maxLen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels=[]\n",
    "for i in model.predict(input_sentence_index):\n",
    "    labels.append(np.argmax(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['I am not feeling good' 'I want to go play some cricket'\n",
      " 'i want to go on a romantic date']\n",
      "[3, 1, 0]\n"
     ]
    }
   ],
   "source": [
    "print input_sentence\n",
    "print labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "I am not feeling good😞\n",
      "I want to go play some cricket⚾\n",
      "i want to go on a romantic date❤\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(input_sentence)):\n",
    "    print input_sentence[i]+emojify_text(str(labels[i]))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}