BalanceNet-1.0.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## BalanceNet prototype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[i] Loaded Parameters:\n",
      " 40000 35 0.2 200 \n",
      " dataset/glove/glove.twitter.27B.200d.txt\n",
      "[i] Importing Modules...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-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"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[+] Using Keras version 2.1.5\n",
      "[+] Finished Importing Modules\n",
      "[i] Reading from csv file...Done!\n",
      "[i] Found 34359 unique tokens.\n",
      "[+] Shape of data tensor: (47288, 30)\n",
      "[+] Shape of label tensor: (47288, 5)\n",
      "[i] Number of entries in each category:\n",
      "[+] Training:\n",
      " [ 7683. 12945. 12834.  3477.   892.]\n",
      "[+] Validation:\n",
      " [1960. 3352. 3104.  824.  217.]\n",
      "[i] Loading GloVe from: dataset/glove/glove.twitter.27B.200d.txt ...Done.\n",
      "[+] Proceeding with Embedding Matrix...[i] Completed!\n",
      "[i] Finished running setup.\n"
     ]
    }
   ],
   "source": [
    "%run Setup.ipynb\n",
    "%run ExtraFunctions.ipynb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Completed!\n"
     ]
    }
   ],
   "source": [
    "# second embedding matrix for non-static channel\n",
    "embedding_matrix_ns = np.random.random((len(word_index) + 1, EMBEDDING_DIM))\n",
    "for word, i in word_index.items():\n",
    "    embedding_vector = embeddings_index.get(word)\n",
    "    if embedding_vector is not None:\n",
    "        # words not found in embedding index will be all-zeros.\n",
    "        embedding_matrix_ns[i] = embedding_vector\n",
    "print(\"Completed!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "sequence_input = Input(shape=(MAX_SEQUENCE_LENGTH,), dtype='int32')\n",
    "\n",
    "# static channel\n",
    "embedding_layer_frozen = Embedding(len(word_index) + 1,\n",
    "                            EMBEDDING_DIM,\n",
    "                            weights=[embedding_matrix],\n",
    "                            input_length=MAX_SEQUENCE_LENGTH,\n",
    "                            trainable=False)\n",
    "embedded_sequences_frozen = embedding_layer_frozen(sequence_input)\n",
    "\n",
    "# non-static channel\n",
    "embedding_layer_train = Embedding(len(word_index) + 1,\n",
    "                            EMBEDDING_DIM,\n",
    "                            weights=[embedding_matrix_ns],\n",
    "                            input_length=MAX_SEQUENCE_LENGTH,\n",
    "                            trainable=True)\n",
    "embedded_sequences_train = embedding_layer_train(sequence_input)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First Half: LSTM > CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_lstm1f = Bidirectional(LSTM(6,return_sequences=True,dropout=0.3, recurrent_dropout=0.0))(embedded_sequences_frozen)\n",
    "l_lstm1t = Bidirectional(LSTM(6,return_sequences=True,dropout=0.3, recurrent_dropout=0.0))(embedded_sequences_train)\n",
    "l_lstm1 = Concatenate(axis=1)([l_lstm1f, l_lstm1t])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/python/util/deprecation.py:497: calling conv1d (from tensorflow.python.ops.nn_ops) with data_format=NHWC is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "`NHWC` for data_format is deprecated, use `NWC` instead\n"
     ]
    }
   ],
   "source": [
    "l_conv_2 = Conv1D(filters=24,kernel_size=2,activation='relu')(l_lstm1)\n",
    "l_conv_2 = Dropout(0.3)(l_conv_2)\n",
    "l_conv_3 = Conv1D(filters=24,kernel_size=3,activation='relu')(l_lstm1)\n",
    "l_conv_3 = Dropout(0.3)(l_conv_3)\n",
    "\n",
    "l_conv_5 = Conv1D(filters=24,kernel_size=5,activation='relu',)(l_lstm1)\n",
    "l_conv_5 = Dropout(0.3)(l_conv_5)\n",
    "l_conv_6 = Conv1D(filters=24,kernel_size=6,activation='relu',kernel_regularizer=regularizers.l2(0.0001))(l_lstm1)\n",
    "l_conv_6 = Dropout(0.3)(l_conv_6)\n",
    "\n",
    "l_conv_8 = Conv1D(filters=24,kernel_size=8,activation='relu',kernel_regularizer=regularizers.l2(0.0001))(l_lstm1)\n",
    "l_conv_8 = Dropout(0.3)(l_conv_8)\n",
    "\n",
    "conv_1 = [l_conv_6,l_conv_5, l_conv_8,l_conv_2,l_conv_3]\n",
    "\n",
    "l_lstm_c = Concatenate(axis=1)(conv_1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Second Half: CNN > LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_conv_4f = Conv1D(filters=12,kernel_size=4,activation='relu',kernel_regularizer=regularizers.l2(0.0001))(embedded_sequences_frozen)\n",
    "l_conv_4f = Dropout(0.3)(l_conv_4f)\n",
    "l_conv_4t = Conv1D(filters=12,kernel_size=4,activation='relu',kernel_regularizer=regularizers.l2(0.0001))(embedded_sequences_train)\n",
    "l_conv_4t = Dropout(0.3)(l_conv_4t)\n",
    "\n",
    "l_conv_3f = Conv1D(filters=12,kernel_size=3,activation='relu',)(embedded_sequences_frozen)\n",
    "l_conv_3f = Dropout(0.3)(l_conv_3f)\n",
    "l_conv_3t = Conv1D(filters=12,kernel_size=3,activation='relu',)(embedded_sequences_train)\n",
    "l_conv_3t = Dropout(0.3)(l_conv_3t)\n",
    "\n",
    "l_conv_2f = Conv1D(filters=12,kernel_size=2,activation='relu')(embedded_sequences_frozen)\n",
    "l_conv_2f = Dropout(0.3)(l_conv_2f)\n",
    "l_conv_2t = Conv1D(filters=12,kernel_size=2,activation='relu')(embedded_sequences_train)\n",
    "l_conv_2t = Dropout(0.3)(l_conv_2t)\n",
    "\n",
    "conv_2 = [l_conv_4f, l_conv_4t,l_conv_3f, l_conv_3t, l_conv_2f, l_conv_2t]\n",
    "\n",
    "l_merge_2 = Concatenate(axis=1)(conv_2)\n",
    "l_c_lstm = Bidirectional(LSTM(12,return_sequences=True,dropout=0.3, recurrent_dropout=0.0))(l_merge_2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Combine both halfs of the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_merge = Concatenate(axis=1)([l_lstm_c, l_c_lstm])\n",
    "l_pool = MaxPooling1D(4)(l_merge)\n",
    "l_drop = Dropout(0.5)(l_pool)\n",
    "l_flat = Flatten()(l_drop)\n",
    "l_dense = Dense(26, activation='relu')(l_flat)\n",
    "preds = Dense(5, activation='softmax')(l_dense)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Model(sequence_input, preds)\n",
    "adadelta = optimizers.Adadelta(lr=0.9, rho=0.95, epsilon=None, decay=0.002)\n",
    "lr_metric = get_lr_metric(adadelta)\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer=adadelta,\n",
    "              metrics=['acc'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rm: logs: No such file or directory\r\n"
     ]
    }
   ],
   "source": [
    "!rm -r logs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "tensorboard = callbacks.TensorBoard(log_dir='./logs', histogram_freq=0, batch_size=16, write_grads=True , write_graph=True)\n",
    "model_checkpoints = callbacks.ModelCheckpoint(\"checkpoint-{val_loss:.3f}.h5\", monitor='val_loss', verbose=0, save_best_only=True, save_weights_only=False, mode='auto', period=0)\n",
    "lr_schedule = callbacks.LearningRateScheduler(initial_boost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 30)           0                                            \n",
      "__________________________________________________________________________________________________\n",
      "embedding_1 (Embedding)         (None, 30, 200)      6872000     input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "embedding_2 (Embedding)         (None, 30, 200)      6872000     input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_1 (Bidirectional) (None, 30, 12)       9936        embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_2 (Bidirectional) (None, 30, 12)       9936        embedding_2[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 60, 12)       0           bidirectional_1[0][0]            \n",
      "                                                                 bidirectional_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_6 (Conv1D)               (None, 27, 12)       9612        embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_7 (Conv1D)               (None, 27, 12)       9612        embedding_2[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_8 (Conv1D)               (None, 28, 12)       7212        embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_9 (Conv1D)               (None, 28, 12)       7212        embedding_2[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_10 (Conv1D)              (None, 29, 12)       4812        embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_11 (Conv1D)              (None, 29, 12)       4812        embedding_2[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_4 (Conv1D)               (None, 55, 24)       1752        concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_3 (Conv1D)               (None, 56, 24)       1464        concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_5 (Conv1D)               (None, 53, 24)       2328        concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_1 (Conv1D)               (None, 59, 24)       600         concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_2 (Conv1D)               (None, 58, 24)       888         concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_6 (Dropout)             (None, 27, 12)       0           conv1d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_7 (Dropout)             (None, 27, 12)       0           conv1d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_8 (Dropout)             (None, 28, 12)       0           conv1d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_9 (Dropout)             (None, 28, 12)       0           conv1d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_10 (Dropout)            (None, 29, 12)       0           conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_11 (Dropout)            (None, 29, 12)       0           conv1d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_4 (Dropout)             (None, 55, 24)       0           conv1d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_3 (Dropout)             (None, 56, 24)       0           conv1d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_5 (Dropout)             (None, 53, 24)       0           conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 59, 24)       0           conv1d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 58, 24)       0           conv1d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 168, 12)      0           dropout_6[0][0]                  \n",
      "                                                                 dropout_7[0][0]                  \n",
      "                                                                 dropout_8[0][0]                  \n",
      "                                                                 dropout_9[0][0]                  \n",
      "                                                                 dropout_10[0][0]                 \n",
      "                                                                 dropout_11[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 281, 24)      0           dropout_4[0][0]                  \n",
      "                                                                 dropout_3[0][0]                  \n",
      "                                                                 dropout_5[0][0]                  \n",
      "                                                                 dropout_1[0][0]                  \n",
      "                                                                 dropout_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_3 (Bidirectional) (None, 168, 24)      2400        concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 449, 24)      0           concatenate_2[0][0]              \n",
      "                                                                 bidirectional_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_1 (MaxPooling1D)  (None, 112, 24)      0           concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_12 (Dropout)            (None, 112, 24)      0           max_pooling1d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "flatten_1 (Flatten)             (None, 2688)         0           dropout_12[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 26)           69914       flatten_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 5)            135         dense_1[0][0]                    \n",
      "==================================================================================================\n",
      "Total params: 13,886,625\n",
      "Trainable params: 7,014,625\n",
      "Non-trainable params: 6,872,000\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()\n",
    "model.save('BalanceNet.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Variable *= will be deprecated. Use variable.assign_mul if you want assignment to the variable value or 'x = x * y' if you want a new python Tensor object.\n"
     ]
    }
   ],
   "source": [
    "model = keras.models.load_model(\"checkpoint-0.91.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Time to train!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Progress:\n",
      "Train on 37831 samples, validate on 9457 samples\n",
      "Epoch 1/200\n",
      "37831/37831 [==============================] - 203s 5ms/step - loss: 0.8950 - acc: 0.6334 - val_loss: 0.8784 - val_acc: 0.6366\n",
      "Epoch 2/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8755 - acc: 0.6436 - val_loss: 0.8722 - val_acc: 0.6424\n",
      "Epoch 3/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8612 - acc: 0.6486 - val_loss: 0.8691 - val_acc: 0.6449\n",
      "Epoch 4/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8496 - acc: 0.6553 - val_loss: 0.8722 - val_acc: 0.6424\n",
      "Epoch 5/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8428 - acc: 0.6568 - val_loss: 0.8660 - val_acc: 0.6470\n",
      "Epoch 6/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8333 - acc: 0.6618 - val_loss: 0.8743 - val_acc: 0.6426\n",
      "Epoch 7/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8252 - acc: 0.6654 - val_loss: 0.8706 - val_acc: 0.6427\n",
      "Epoch 8/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8189 - acc: 0.6684 - val_loss: 0.8715 - val_acc: 0.6469\n",
      "Epoch 9/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8146 - acc: 0.6702 - val_loss: 0.8727 - val_acc: 0.6457\n",
      "Epoch 10/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8080 - acc: 0.6715 - val_loss: 0.8740 - val_acc: 0.6447\n",
      "Epoch 11/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.8023 - acc: 0.6746 - val_loss: 0.8783 - val_acc: 0.6424\n",
      "Epoch 12/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7980 - acc: 0.6770 - val_loss: 0.8746 - val_acc: 0.6462\n",
      "Epoch 13/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7964 - acc: 0.6769 - val_loss: 0.8767 - val_acc: 0.6414\n",
      "Epoch 14/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7912 - acc: 0.6783 - val_loss: 0.8799 - val_acc: 0.6413\n",
      "Epoch 15/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7873 - acc: 0.6789 - val_loss: 0.8792 - val_acc: 0.6445\n",
      "Epoch 16/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7842 - acc: 0.6831 - val_loss: 0.8811 - val_acc: 0.6413\n",
      "Epoch 17/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7839 - acc: 0.6826 - val_loss: 0.8783 - val_acc: 0.6439\n",
      "Epoch 18/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7788 - acc: 0.6865 - val_loss: 0.8814 - val_acc: 0.6430\n",
      "Epoch 19/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7776 - acc: 0.6874 - val_loss: 0.8828 - val_acc: 0.6446\n",
      "Epoch 20/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7749 - acc: 0.6854 - val_loss: 0.8800 - val_acc: 0.6441\n",
      "Epoch 21/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7716 - acc: 0.6878 - val_loss: 0.8846 - val_acc: 0.6419\n",
      "Epoch 22/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7728 - acc: 0.6863 - val_loss: 0.8830 - val_acc: 0.6431\n",
      "Epoch 23/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7690 - acc: 0.6891 - val_loss: 0.8836 - val_acc: 0.6438\n",
      "Epoch 24/200\n",
      "37831/37831 [==============================] - 195s 5ms/step - loss: 0.7628 - acc: 0.6944 - val_loss: 0.8897 - val_acc: 0.6401\n",
      "Epoch 25/200\n",
      " 6784/37831 [====>.........................] - ETA: 2:31 - loss: 0.7495 - acc: 0.6962"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-30-3ddbc74504ce>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m model_log = model.fit(x_train, y_train, validation_data=(x_val, y_val),\n\u001b[1;32m      3\u001b[0m           \u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m200\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m128\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m           callbacks=[tensorboard, model_checkpoints])\n\u001b[0m\u001b[1;32m      5\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0mpandas\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel_log\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhistory\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto_csv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"history-balance.csv\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)\u001b[0m\n\u001b[1;32m   1710\u001b[0m                               \u001b[0minitial_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitial_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1711\u001b[0m                               \u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1712\u001b[0;31m                               validation_steps=validation_steps)\n\u001b[0m\u001b[1;32m   1713\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1714\u001b[0m     def evaluate(self, x=None, y=None,\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/keras/engine/training.py\u001b[0m in \u001b[0;36m_fit_loop\u001b[0;34m(self, f, ins, out_labels, batch_size, epochs, verbose, callbacks, val_f, val_ins, shuffle, callback_metrics, initial_epoch, steps_per_epoch, validation_steps)\u001b[0m\n\u001b[1;32m   1233\u001b[0m                         \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtoarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1234\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1235\u001b[0;31m                     \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1236\u001b[0m                     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1237\u001b[0m                         \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/keras/backend/tensorflow_backend.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   2473\u001b[0m         \u001b[0msession\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_session\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2474\u001b[0m         updated = session.run(fetches=fetches, feed_dict=feed_dict,\n\u001b[0;32m-> 2475\u001b[0;31m                               **self.session_kwargs)\n\u001b[0m\u001b[1;32m   2476\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mupdated\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2477\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m    893\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    894\u001b[0m       result = self._run(None, fetches, feed_dict, options_ptr,\n\u001b[0;32m--> 895\u001b[0;31m                          run_metadata_ptr)\n\u001b[0m\u001b[1;32m    896\u001b[0m       \u001b[0;32mif\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    897\u001b[0m         \u001b[0mproto_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf_session\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTF_GetBuffer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrun_metadata_ptr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_run\u001b[0;34m(self, handle, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m   1126\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfinal_fetches\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mfinal_targets\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mhandle\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mfeed_dict_tensor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1127\u001b[0m       results = self._do_run(handle, final_targets, final_fetches,\n\u001b[0;32m-> 1128\u001b[0;31m                              feed_dict_tensor, options, run_metadata)\n\u001b[0m\u001b[1;32m   1129\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1130\u001b[0m       \u001b[0mresults\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_run\u001b[0;34m(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m   1342\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mhandle\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1343\u001b[0m       return self._do_call(_run_fn, self._session, feeds, fetches, targets,\n\u001b[0;32m-> 1344\u001b[0;31m                            options, run_metadata)\n\u001b[0m\u001b[1;32m   1345\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1346\u001b[0m       \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_do_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0m_prun_fn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_session\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeeds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetches\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_call\u001b[0;34m(self, fn, *args)\u001b[0m\n\u001b[1;32m   1348\u001b[0m   \u001b[0;32mdef\u001b[0m \u001b[0m_do_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1349\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1350\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1351\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0merrors\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mOpError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1352\u001b[0m       \u001b[0mmessage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcompat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_text\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmessage\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_run_fn\u001b[0;34m(session, feed_dict, fetch_list, target_list, options, run_metadata)\u001b[0m\n\u001b[1;32m   1327\u001b[0m           return tf_session.TF_Run(session, options,\n\u001b[1;32m   1328\u001b[0m                                    \u001b[0mfeed_dict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetch_list\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget_list\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1329\u001b[0;31m                                    status, run_metadata)\n\u001b[0m\u001b[1;32m   1330\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1331\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_prun_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msession\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeed_dict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetch_list\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "print(\"Training Progress:\")\n",
    "model_log = model.fit(x_train, y_train, validation_data=(x_val, y_val),\n",
    "          epochs=200, batch_size=128,\n",
    "          callbacks=[tensorboard, model_checkpoints])\n",
    "\n",
    "pandas.DataFrame(model_log.history).to_csv(\"history-balance.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Testing and Inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import load_model\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "import itertools, pickle\n",
    "\n",
    "with open('tokenizer.pickle', 'rb') as handle:\n",
    "    tokenizer = pickle.load(handle)\n",
    "\n",
    "classes = [\"neutral\", \"happy\", \"sad\", \"hate\",\"anger\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BalanceNet-1.0.ipynb      _config.yml               \u001b[1m\u001b[36mraw_data\u001b[m\u001b[m\r\n",
      "BalanceNet.h5             \u001b[1m\u001b[36marchitectures\u001b[m\u001b[m             \u001b[31mrun_tensorboard.sh\u001b[m\u001b[m\r\n",
      "ClassMerge.ipynb          checkpoint-0.90.h5        \u001b[1m\u001b[36msklearn_experiments\u001b[m\u001b[m\r\n",
      "ExtraFunctions.ipynb      checkpoint-0.91.h5        test_env.py\r\n",
      "ProcessData-Crawled.ipynb data.csv                  tokenizer.pickle\r\n",
      "ProcessData.ipynb         \u001b[1m\u001b[36mdataset\u001b[m\u001b[m                   twitter_crawl.py\r\n",
      "README.md                 \u001b[1m\u001b[36mexperiments\u001b[m\u001b[m\r\n",
      "Setup.ipynb               \u001b[1m\u001b[36mimages\u001b[m\u001b[m\r\n"
     ]
    }
   ],
   "source": [
    "! ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "#model_test = load_model('checkpoint-0.866.h5')\n",
    "model_test = load_model('best_weights.h5')\n",
    "Y_test = np.argmax(y_val, axis=1) # Convert one-hot to index\n",
    "y_pred = model_test.predict(x_val)\n",
    "y_pred_class = np.argmax(y_pred,axis=1)\n",
    "cnf_matrix = confusion_matrix(Y_test, y_pred_class)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "    neutral       0.53      0.39      0.45      1938\n",
      "      happy       0.66      0.74      0.70      3257\n",
      "        sad       0.65      0.70      0.67      3212\n",
      "       hate       0.84      0.72      0.78       840\n",
      "      anger       0.93      0.73      0.82       210\n",
      "\n",
      "avg / total       0.65      0.66      0.65      9457\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(Y_test, y_pred_class, target_names=classes))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f235d009a58>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 725,
       "width": 839
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_confusion_matrix(cm, labels,\n",
    "                          normalize=True,\n",
    "                          title='Confusion Matrix (Validation Set)',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        #print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        #print('Confusion matrix, without normalization')\n",
    "        pass\n",
    "\n",
    "    #print(cm)\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(labels))\n",
    "    plt.xticks(tick_marks, labels, rotation=45)\n",
    "    plt.yticks(tick_marks, labels)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "\n",
    "plt.figure(figsize=(20,10))\n",
    "plot_confusion_matrix(cnf_matrix, labels=classes)\n",
    "\n",
    "# precision = true_pos / (true_pos + false_pos)\n",
    "# recall = true_pos / (true_pos + false_neg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "text = [\"I salute you for the bravery and sacrifice! A true hero indeed.\",\n",
    "        \"I am sorry but I trust HRW I damned sight more than the PAP and it's cronies! Off course the PAP will say that they (HRW) made things up...despite of the fact that SG is a dictator state!?\",\n",
    "        \"PAP are taking the piss again!\",\n",
    "        \"Thought he sold his kidney to buy it; Instead, he bought a kidney then bought the car Filthy rich This is why we need communism\",\n",
    "        \"Somebody needs to water Tharman's head, hair needs to be grown there\",\n",
    "        \"what a nuisance fk. a proper clean and flat footpath,,now obstructed by sharedbikes..! which idiotic MP allowed this to happen?\",\n",
    "        \"What baby bonus scheme ??? To grow up a kid in Singapore you think is easy now bo ??? Both parent need to work to grow up a kid until 21 , you think tats easy bo ??? Think la\"\n",
    "       ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "I salute you for the bravery and sacrifice! A true hero indeed. \n",
      "Prediction: happy \n",
      "\n",
      "I am sorry but I trust HRW I damned sight more than the PAP and it's cronies! Off course the PAP will say that they (HRW) made things up...despite of the fact that SG is a dictator state!? \n",
      "Prediction: sad \n",
      "\n",
      "PAP are taking the piss again! \n",
      "Prediction: sad \n",
      "\n",
      "Thought he sold his kidney to buy it; Instead, he bought a kidney then bought the car Filthy rich This is why we need communism \n",
      "Prediction: hate \n",
      "\n",
      "Somebody needs to water Tharman's head, hair needs to be grown there \n",
      "Prediction: sad \n",
      "\n",
      "what a nuisance fk. a proper clean and flat footpath,,now obstructed by sharedbikes..! which idiotic MP allowed this to happen? \n",
      "Prediction: sad \n",
      "\n",
      "What baby bonus scheme ??? To grow up a kid in Singapore you think is easy now bo ??? Both parent need to work to grow up a kid until 21 , you think tats easy bo ??? Think la \n",
      "Prediction: sad \n",
      "\n"
     ]
    }
   ],
   "source": [
    "sequences_test = tokenizer.texts_to_sequences(text)\n",
    "data_int_t = pad_sequences(sequences_test, padding='pre', maxlen=(MAX_SEQUENCE_LENGTH-5))\n",
    "data_test = pad_sequences(data_int_t, padding='post', maxlen=(MAX_SEQUENCE_LENGTH))\n",
    "y_prob = model.predict(data_test)\n",
    "for n, prediction in enumerate(y_prob):\n",
    "    pred = y_prob.argmax(axis=-1)[n]\n",
    "    print(text[n],\"\\nPrediction:\",classes[pred],\"\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([   1, 6097,    5,   12,    3,    6, 4841,    4,  515, 1476, 1238])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(sequences_test[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}