CNN_GRU_sentiment.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN_GRU_sentiment.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "36ZYOuGsQWOq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Load the Drive helper and mount\n",
        "from google.colab import drive"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V1OoZjWRQeE1",
        "colab_type": "code",
        "outputId": "8b8ac712-74ad-4dc5-a1bd-23f63f006676",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "!pwd\n",
        "%cd /content/drive/My Drive/Term5/NLP/Project"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\n",
            "/content/drive/My Drive/Term5/NLP/Project\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PcH74oGLXNNO",
        "colab_type": "code",
        "outputId": "099a520f-72ee-430b-8af9-6f81e2e9863a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import pandas as pd\n",
        "from keras.preprocessing.text import Tokenizer"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p5DJ4gCaXDTg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def readFile(fileName):\n",
        "  filedata = pd.read_csv(fileName, delimiter='\\t', header = None, skiprows=1)\n",
        "  filedata.columns = [\"id\", \"drugName\", \"condition\", \"review\", \"rating\", \"date\", \"usefulCount\", \"clean_review\", \"sentiment\"]\n",
        "  filedata.drop(['review'], axis=1, inplace = True)\n",
        "  filedata.dropna(axis=0, inplace=True)\n",
        "  return filedata"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_MBVvilSQeqI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "dataPath = '/content/drive/My Drive/Term5/NLP/Project/Dataset'\n",
        "outputPath = '/content/drive/My Drive/Term5/NLP/Project/Model_output/CNN_GRU/'\n",
        "glove_vectors_file = dataPath + '/../word_embeddings/glove.840B.300d.txt'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3QC83lC8W5Ex",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "drug_train = dataPath + '/cleaned_Train_data.csv'\n",
        "drug_test = dataPath + '/cleaned_Test_data.csv'\n",
        "drug_valid = dataPath + '/cleaned_Valid_data.csv'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l1N8F32cW928",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_data = readFile(drug_train)\n",
        "valid_data = readFile(drug_valid)\n",
        "test_data = readFile(drug_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "63_uF3AWYxSL",
        "colab_type": "code",
        "outputId": "537914b1-cb0c-4555-b8cb-b42e34ac81db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 141
        }
      },
      "source": [
        "train_data.head(3)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>drugName</th>\n",
              "      <th>condition</th>\n",
              "      <th>rating</th>\n",
              "      <th>date</th>\n",
              "      <th>usefulCount</th>\n",
              "      <th>clean_review</th>\n",
              "      <th>sentiment</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>204048</td>\n",
              "      <td>Xerac AC</td>\n",
              "      <td>Hyperhidrosis</td>\n",
              "      <td>10.0</td>\n",
              "      <td>May 15, 2010</td>\n",
              "      <td>10</td>\n",
              "      <td>wow struggl excess underarm sweat sever year r...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>52764</td>\n",
              "      <td>Carbidopa / levodopa</td>\n",
              "      <td>Parkinson's Disease</td>\n",
              "      <td>4.0</td>\n",
              "      <td>December 30, 2016</td>\n",
              "      <td>17</td>\n",
              "      <td>rytari month first work well pill mg mg time d...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>210591</td>\n",
              "      <td>Mirtazapine</td>\n",
              "      <td>Post Traumatic Stress Disorde</td>\n",
              "      <td>7.0</td>\n",
              "      <td>September 1, 2017</td>\n",
              "      <td>6</td>\n",
              "      <td>hey least one make manic suicid like zoloft pr...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       id  ... sentiment\n",
              "0  204048  ...         1\n",
              "1   52764  ...         0\n",
              "2  210591  ...         1\n",
              "\n",
              "[3 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uc_0I3LNQOEz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import io\n",
        "import numpy as np"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YMvspDcaRY2R",
        "colab_type": "code",
        "outputId": "35eeb086-646d-4d66-b399-7f582d2063d0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "#http://nlp.stanford.edu/data/glove.840B.300d.zip\n",
        "glove_embeddings = {}\n",
        "\n",
        "print('Loading '+ glove_vectors_file)\n",
        "with open(glove_vectors_file, 'r') as f:\n",
        "  for line in f:\n",
        "    values = line.split(' ')\n",
        "    word = values[0]\n",
        "    embedding = np.asarray(values[1:], dtype='float32')\n",
        "    glove_embeddings[word] = embedding\n",
        "        \n",
        "print('Embeddings size: %d' % len(glove_embeddings))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Loading /content/drive/My Drive/Term5/NLP/Project/Dataset/../word_embeddings/glove.840B.300d.txt\n",
            "Embeddings size: 2196016\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "06vLHnysReAD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embedding_dim = 300"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NcjTJusKRexy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def convertSentencesToList(traindata, validdata):\n",
        "  train_sentence = traindata['clean_review'].values.tolist()\n",
        "  valid_sentence = validdata['clean_review'].values.tolist()\n",
        "  return train_sentence, valid_sentence"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B4zqDEVHSzKk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def converToWordSequences(embeddinglength, training_data, t_s, v_s):\n",
        "  tokenizer = Tokenizer(num_words = embeddinglength)\n",
        "  tokenizer.fit_on_texts(training_data)\n",
        "\n",
        "  t_s_word_sequences = tokenizer.texts_to_sequences(t_s)\n",
        "\n",
        "  v_s_word_sequences = tokenizer.texts_to_sequences(v_s)\n",
        "  return tokenizer, t_s_word_sequences, v_s_word_sequences"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RPTzNdm0TREF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_sent, valid_sent = convertSentencesToList(train_data, valid_data)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JuNmdPgKTbLe",
        "colab_type": "code",
        "outputId": "b8a392c7-c467-4e2c-8233-f1052292664c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "glove_length = len(glove_embeddings)\n",
        "\n",
        "print('Found '+ str(len(train_sent)) + ' in training sentence')\n",
        "\n",
        "training_sent = train_sent\n",
        "valid_sent = valid_sent\n",
        "\n",
        "tokenizer_g, t_s_word_sequences_g, v_s_word_sequences_g = converToWordSequences(glove_length, training_sent, train_sent, valid_sent)\n",
        "word_index_g = tokenizer_g.word_index\n",
        "\n",
        "print('Found '+str(len(word_index_g))+' unique tokens for glove sequencing.')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 127122 in training sentence\n",
            "Found 29822 unique tokens for glove sequencing.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4B1GxZwETyxj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from collections import defaultdict"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SlAUgGggTzjx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def useWordEmbeddings(word_index, NUM_WORDS, embeddings, sentence_word_sequences):\n",
        "  words_len = min(NUM_WORDS, len(word_index))\n",
        "  # word_embedding_matrix = np.zeros((words_len + 1, embedding_dim))\n",
        "  word_embedding_matrix = np.random.random((words_len + 1, embedding_dim))\n",
        "  k = 0\n",
        "  for word, i in word_index.items():\n",
        "    if i >= NUM_WORDS:\n",
        "      continue\n",
        "    embedding_vector = embeddings.get(word)\n",
        "    if embedding_vector is not None:\n",
        "      word_embedding_matrix[i] = embedding_vector\n",
        "      k += 1\n",
        "\n",
        "  return word_embedding_matrix"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9_1aboY5Um-z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "word_embedding_matrix_g = useWordEmbeddings(word_index_g, glove_length, glove_embeddings, t_s_word_sequences_g)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gdxssUK_VAQj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "max_seq_length = 300"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W59j9a5CVBh7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.preprocessing.sequence import pad_sequences"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LByN8bABVFnl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def convertToPadSequences(sentence_word_sequences, valid_sentence_word_sequences, max_seq_length):\n",
        "  s1_data = pad_sequences(sentence_word_sequences, maxlen = max_seq_length)\n",
        "  s1_dataValid = pad_sequences(valid_sentence_word_sequences, maxlen = max_seq_length)\n",
        "  return s1_data, s1_dataValid"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PE6xd1LlVLRz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def printSentenceInfo(s1_data, labels):\n",
        "  print('Shape of sentence1 tensor:', s1_data.shape)\n",
        "  print('Shape of label tensor:', labels.shape)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aPe_k-Y0VLvv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "s1_data_g, s1_dataValid_g = convertToPadSequences(t_s_word_sequences_g, v_s_word_sequences_g, max_seq_length)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-DiioFVKezVo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.utils import np_utils, plot_model\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "le = LabelEncoder()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hAPy-bzRfbGb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def convertToCategorical(data):\n",
        "  labels = np_utils.to_categorical(le.fit_transform(data[\"sentiment\"].values)).astype(\"int64\")\n",
        "  return labels"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "--LiUorfVYBh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_labels = convertToCategorical(train_data)\n",
        "valid_labels = convertToCategorical(valid_data)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ngw8VJvhVYtE",
        "colab_type": "code",
        "outputId": "7879f9b6-9231-4d19-dd25-da82d6f75fab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "printSentenceInfo(s1_data_g, train_labels)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shape of sentence1 tensor: (127122, 300)\n",
            "Shape of label tensor: (127122, 2)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fiEvFcqfVazo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.layers import LSTM, Embedding, concatenate, Dense, PReLU, Dropout, Activation, Conv1D, MaxPooling1D, Flatten, SimpleRNN, GRU\n",
        "from keras.models import Sequential\n",
        "from keras.initializers import Constant\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "from keras.models import Model\n",
        "from keras.optimizers import Adam\n",
        "from keras.callbacks import ModelCheckpoint\n",
        "from IPython.display import Image\n",
        "import time\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y3woEneNVfCZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "num_rnn = np.random.randint(175, 275)\n",
        "num_dense = np.random.randint(100, 150)\n",
        "rate_drop_rnn = 0.15 + np.random.rand() * 0.25\n",
        "rate_drop_dense = 0.15 + np.random.rand() * 0.25"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NHPGKygMgAMw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "batch_size = 128"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dz8wDqmugDM_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def trainModel(model, model_name, s1_data, labels, batch_size, s1_data_valid, labels_valid):\n",
        "  checkpoint = ModelCheckpoint(outputPath+ model_name + '-checkpoint-weights.{epoch:02d}-{val_accuracy:.2f}.hdf5', monitor='val_accuracy', save_best_only=True)\n",
        "  history = model.fit([s1_data],\n",
        "                        labels,\n",
        "                        epochs = 10,\n",
        "                        batch_size = batch_size,\n",
        "                        validation_data=([s1_data_valid], labels_valid),\n",
        "                        shuffle = False,\n",
        "                        verbose = 2,\n",
        "                        callbacks = [checkpoint])\n",
        "  return history"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1NPQ7EPkgF5Y",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plotModelPicture(model, model_name):\n",
        "  plot_model(model, to_file = outputPath + model_name + '-model.png', show_shapes=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8l7bXjfqgGZ4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plotMetricsGraph(modelhistory, model_name):\n",
        "  # summarize history for accuracy\n",
        "  plt.plot(modelhistory.history['accuracy'])\n",
        "  plt.plot(modelhistory.history['val_accuracy'])\n",
        "  plt.title('model accuracy')\n",
        "  plt.ylabel('accuracy')\n",
        "  plt.xlabel('epoch')\n",
        "  plt.legend(['train', 'valid'], loc='upper left')\n",
        "  plt.show()\n",
        "  plt.savefig(model_name + '-acc.png', dpi = 300)\n",
        "\n",
        "  # summarize history for loss\n",
        "  plt.plot(modelhistory.history['loss'])\n",
        "  plt.plot(modelhistory.history['val_loss'])\n",
        "  plt.title('model loss')\n",
        "  plt.ylabel('loss')\n",
        "  plt.xlabel('epoch')\n",
        "  plt.legend(['train', 'valid'], loc='upper left')\n",
        "  plt.show()\n",
        "  plt.savefig(model_name + '-loss.png', dpi = 300)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zcGsRyCUEsKk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def evaluateModel(model, test_s_data, test_labels):\n",
        "  loss, accuracy = model.evaluate([test_s_data], test_labels)\n",
        "  print('loss     = ', loss)\n",
        "  print('accuracy = ', accuracy)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nEh3hj_DCmHn",
        "colab_type": "code",
        "outputId": "bb2e7965-27e3-47a0-d140-458c8e78be0d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def plotConfusionMatrix(test_y, y_pred):\n",
        "  #y_pred=model.predict(test_x,batch_size=15)\n",
        "  #cm =confusion_matrix(test_y.argmax(axis=1), y_pred.argmax(axis=1))\n",
        "  cm =confusion_matrix(test_y, y_pred)\n",
        "  print(cm)\n",
        "  index = ['Negative','Positive']\n",
        "  columns = ['Negative','Positive']\n",
        "  cm_df = pd.DataFrame(cm,columns,index)             \n",
        "  plt.figure(figsize=(10,6))\n",
        "  sns.heatmap(cm_df, annot=True)\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LyzwKC5d-EiN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def CNNGRUModel(word_index, EMBEDDING_DIM, embedding_matrix, MAX_SEQUENCE_LENGTH, p_drop=0.0):\n",
        "  encoder_1 = Sequential()\n",
        "  encoder_1.add(Embedding(len(word_index) + 1, EMBEDDING_DIM,\n",
        "                                    embeddings_initializer=Constant(embedding_matrix),\n",
        "                                    input_length=MAX_SEQUENCE_LENGTH,\n",
        "                                    trainable=False))\n",
        "\n",
        "  encoder_1.add(Conv1D(128, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding=\"same\"))\n",
        "  encoder_1.add(Conv1D(64, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(64, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(32, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(32, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(16, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(16, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(8, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(Conv1D(8, kernel_size = 2, padding = \"same\"))\n",
        "  encoder_1.add(MaxPooling1D(pool_size=2, padding='same'))\n",
        "  encoder_1.add(GRU(128,return_sequences=True))\n",
        "  encoder_1.add(Dropout(0.3))\n",
        "  encoder_1.add(Flatten())\n",
        "  encoder_1.add(Dense(128,activation='relu'))\n",
        "  encoder_1.add(Dropout(0.5))\n",
        "  encoder_1.add(Dense(2, activation=\"sigmoid\", name=\"final_output\"))\n",
        "\n",
        "  encoder_1.compile(loss = 'categorical_crossentropy', optimizer=Adam(lr=0.001), metrics = ['accuracy'])\n",
        "    \n",
        "  return encoder_1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8sb0LRoohEH4",
        "colab_type": "code",
        "outputId": "6050cd54-ce6c-4f08-86b5-3f4046ccb761",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "model_cnn_gru_g = CNNGRUModel(word_index_g, embedding_dim, word_embedding_matrix_g, max_seq_length, p_drop=0.25)\n",
        "model_cnn_gru_g.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_3 (Embedding)      (None, 300, 300)          8946900   \n",
            "_________________________________________________________________\n",
            "conv1d_19 (Conv1D)           (None, 300, 128)          76928     \n",
            "_________________________________________________________________\n",
            "max_pooling1d_19 (MaxPooling (None, 150, 128)          0         \n",
            "_________________________________________________________________\n",
            "conv1d_20 (Conv1D)           (None, 150, 64)           16448     \n",
            "_________________________________________________________________\n",
            "max_pooling1d_20 (MaxPooling (None, 75, 64)            0         \n",
            "_________________________________________________________________\n",
            "conv1d_21 (Conv1D)           (None, 75, 64)            8256      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_21 (MaxPooling (None, 38, 64)            0         \n",
            "_________________________________________________________________\n",
            "conv1d_22 (Conv1D)           (None, 38, 32)            4128      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_22 (MaxPooling (None, 19, 32)            0         \n",
            "_________________________________________________________________\n",
            "conv1d_23 (Conv1D)           (None, 19, 32)            2080      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_23 (MaxPooling (None, 10, 32)            0         \n",
            "_________________________________________________________________\n",
            "conv1d_24 (Conv1D)           (None, 10, 16)            1040      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_24 (MaxPooling (None, 5, 16)             0         \n",
            "_________________________________________________________________\n",
            "conv1d_25 (Conv1D)           (None, 5, 16)             528       \n",
            "_________________________________________________________________\n",
            "max_pooling1d_25 (MaxPooling (None, 3, 16)             0         \n",
            "_________________________________________________________________\n",
            "conv1d_26 (Conv1D)           (None, 3, 8)              264       \n",
            "_________________________________________________________________\n",
            "max_pooling1d_26 (MaxPooling (None, 2, 8)              0         \n",
            "_________________________________________________________________\n",
            "conv1d_27 (Conv1D)           (None, 2, 8)              136       \n",
            "_________________________________________________________________\n",
            "max_pooling1d_27 (MaxPooling (None, 1, 8)              0         \n",
            "_________________________________________________________________\n",
            "gru_2 (GRU)                  (None, 1, 128)            52608     \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 1, 128)            0         \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 128)               16512     \n",
            "_________________________________________________________________\n",
            "dropout_4 (Dropout)          (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "final_output (Dense)         (None, 2)                 258       \n",
            "=================================================================\n",
            "Total params: 9,126,086\n",
            "Trainable params: 179,186\n",
            "Non-trainable params: 8,946,900\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YPa-pu43hohU",
        "colab_type": "code",
        "outputId": "125b811e-f916-487b-8dd6-7e08df46dcc3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "model_name = 'CNN_GRU_Glove'\n",
        "plotModelPicture(model_cnn_gru_g, model_name)\n",
        "Image(filename = outputPath + model_name + '-model.png')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VeIxQDj68L5Z",
        "colab_type": "code",
        "outputId": "3c7bbbd9-2a80-41cf-8437-3cd9f238727a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 377
        }
      },
      "source": [
        "history_cnn_gru_g = trainModel(model_cnn_gru_g, model_name, s1_data_g, train_labels, batch_size, s1_dataValid_g, valid_labels)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 127122 samples, validate on 31607 samples\n",
            "Epoch 1/10\n",
            " - 363s - loss: 0.4719 - accuracy: 0.7787 - val_loss: 0.4066 - val_accuracy: 0.8186\n",
            "Epoch 2/10\n",
            " - 355s - loss: 0.3824 - accuracy: 0.8331 - val_loss: 0.3838 - val_accuracy: 0.8319\n",
            "Epoch 3/10\n",
            " - 349s - loss: 0.3350 - accuracy: 0.8587 - val_loss: 0.3738 - val_accuracy: 0.8400\n",
            "Epoch 4/10\n",
            " - 346s - loss: 0.2954 - accuracy: 0.8795 - val_loss: 0.3768 - val_accuracy: 0.8430\n",
            "Epoch 5/10\n",
            " - 352s - loss: 0.2617 - accuracy: 0.8961 - val_loss: 0.3825 - val_accuracy: 0.8489\n",
            "Epoch 6/10\n",
            " - 350s - loss: 0.2345 - accuracy: 0.9085 - val_loss: 0.4566 - val_accuracy: 0.8394\n",
            "Epoch 7/10\n",
            " - 353s - loss: 0.2103 - accuracy: 0.9195 - val_loss: 0.5091 - val_accuracy: 0.8318\n",
            "Epoch 8/10\n",
            " - 356s - loss: 0.1918 - accuracy: 0.9278 - val_loss: 0.4154 - val_accuracy: 0.8544\n",
            "Epoch 9/10\n",
            " - 350s - loss: 0.1782 - accuracy: 0.9331 - val_loss: 0.4022 - val_accuracy: 0.8606\n",
            "Epoch 10/10\n",
            " - 343s - loss: 0.1589 - accuracy: 0.9408 - val_loss: 0.4276 - val_accuracy: 0.8613\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IYhyWc2w-IEC",
        "colab_type": "code",
        "outputId": "dd77e547-c6c2-4d68-b198-a8cff41a223c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 590
        }
      },
      "source": [
        "plotMetricsGraph(history_cnn_gru_g, model_name)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hGn7R8iKDspw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def getTestDataSequences(tokenizer):\n",
        "  test_s_word_sequences = tokenizer.texts_to_sequences(test_data['clean_review'].values.tolist())\n",
        "\n",
        "  test_s_data = pad_sequences(test_s_word_sequences, maxlen = max_seq_length)\n",
        "  test_labels = np_utils.to_categorical(le.fit_transform(test_data[\"sentiment\"].values)).astype(\"int64\")\n",
        "\n",
        "  print('Shape of test sentence1 tensor:', test_s_data.shape)\n",
        "  print('Shape of test label tensor:', test_labels.shape)\n",
        "  return test_s_data, test_labels"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_EBxNREdv9KK",
        "colab_type": "code",
        "outputId": "13b62dd2-e67d-4913-de14-be9ecd6daf7e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "test_s_data_g, test_labels_g = getTestDataSequences(tokenizer_g)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shape of test sentence1 tensor: (52856, 300)\n",
            "Shape of test label tensor: (52856, 2)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aMuAHTZ_v_b3",
        "colab_type": "code",
        "outputId": "c37b9a3e-9590-4cf1-949a-e3e335d834a1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "evaluateModel(model_cnn_gru_g, test_s_data_g, test_labels_g)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "52856/52856 [==============================] - 54s 1ms/step\n",
            "loss     =  0.4372151943536277\n",
            "accuracy =  0.8656402277946472\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BbO6otG7wCqy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = model_cnn_gru_g"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1Id-vzpwwFeQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_pred = model.predict([test_s_data_g], batch_size=32)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OfDcf-f0wH6S",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "output = list()\n",
        "output_score = list()\n",
        "for x in test_pred:\n",
        "  #print(type(x))\n",
        "  output_score.append(np.amax(x))\n",
        "  output.append(np.argmax(x))\n",
        "\n",
        "test_data[\"Actual Output\"] = output\n",
        "test_data[\"Predicted Score\"] = output_score\n",
        "\n",
        "test_data.to_excel(outputPath+'output_cnn_gru_glove.xlsx')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0B_0DmjxwOvO",
        "colab_type": "code",
        "outputId": "6def8c75-586a-4cf8-eb96-864081c2a53a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 171
        }
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "\n",
        "predicted = np.argmax(test_pred, axis=1)\n",
        "report = classification_report(test_data[\"sentiment\"], test_data[\"Actual Output\"])\n",
        "print(report)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.77      0.77     15962\n",
            "           1       0.90      0.90      0.90     36894\n",
            "\n",
            "    accuracy                           0.87     52856\n",
            "   macro avg       0.83      0.83      0.83     52856\n",
            "weighted avg       0.86      0.86      0.86     52856\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LrfJo7KWwQ1w",
        "colab_type": "code",
        "outputId": "a9195c76-2f40-4c20-f3b8-d0eabdef3e8e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 408
        }
      },
      "source": [
        "plotConfusionMatrix(test_data[\"sentiment\"], test_data[\"Actual Output\"])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[12278  3684]\n",
            " [ 3682 33212]]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}