Improved_LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "9e5d0f00",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sklearn\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import csv, re\n",
    "import string\n",
    "import codecs\n",
    "import nltk\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "import sklearn.metrics as metrics\n",
    "import matplotlib.pyplot as plt\n",
    "import jsonlines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "1a72516b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data into a pandas dataframe\n",
    "\n",
    "data = pd.read_csv('/data/wcpr_mypersonality.csv',encoding='mac-roman')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "b946b001",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['#AUTHID', 'STATUS', 'sEXT', 'sNEU', 'sAGR', 'sCON', 'sOPN', 'cEXT',\n",
       "       'cNEU', 'cAGR', 'cCON', 'cOPN', 'DATE', 'NETWORKSIZE', 'BETWEENNESS',\n",
       "       'NBETWEENNESS', 'DENSITY', 'BROKERAGE', 'NBROKERAGE', 'TRANSITIVITY'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "6701dea6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert labels to numbers\n",
    "data.replace({'y':1,'n':0}, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "id": "20f017a4",
   "metadata": {},
   "outputs": [
    {
     "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>Status</th>\n",
       "      <th>Target Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>likes the sound of thunder.</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>is so sleepy it's not even funny that's she ca...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>is sore and wants the knot of muscles at the b...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>likes how the day sounds in this new song.</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>is home. &lt;3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              Status  Target Label\n",
       "0                        likes the sound of thunder.             1\n",
       "1  is so sleepy it's not even funny that's she ca...             1\n",
       "2  is sore and wants the knot of muscles at the b...             1\n",
       "3         likes how the day sounds in this new song.             1\n",
       "4                                        is home. <3             1"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Status and neurotic target label columns\n",
    "data = pd.DataFrame({'Status': data.STATUS, 'Target Label': data.cNEU})\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "id": "138a1d91",
   "metadata": {},
   "outputs": [],
   "source": [
    "# assign values\n",
    "status = data['Status'].values\n",
    "labels = data['Target Label'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "id": "142dc291",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9917, 9917)"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check number of rows in status and labels\n",
    "len(labels),len(status)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "id": "49f18f0c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, \"is so sleepy it's not even funny that's she can't get to sleep.\")"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sample status and label\n",
    "labels[1],status[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "40b992b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# fix random seed for reproducibility\n",
    "tf.random.set_seed(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "id": "070b355b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# split X and y training and testing(validate) datasets (same 80-20 as basemodel)\n",
    "training_portion = 0.7\n",
    "train_size = int(len(status) * training_portion)\n",
    "\n",
    "train_dataset = status[0: train_size]\n",
    "train_labels = labels[0: train_size]\n",
    "\n",
    "test_dataset = status[train_size:]\n",
    "validation_labels = labels[train_size:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "id": "db7009fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load the dataset but only keep the top n words, zero the rest\n",
    "top_words = 5000\n",
    "vocab_size = 5000\n",
    "embedding_dim = 32 #Embedding size for each token \n",
    "max_length = 200\n",
    "oov_tok = '<OOV>' #  Out of Vocabulary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "id": "a1cc6dc0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train_status 6941\n",
      "train_labels 6941\n",
      "validation_status 2976\n",
      "validation_labels 2976\n"
     ]
    }
   ],
   "source": [
    "# output train and test sizes\n",
    "print('train_status' ,len(train_dataset))\n",
    "print('train_labels', len(train_labels))\n",
    "print('validation_status', len(test_dataset))\n",
    "print('validation_labels', len(validation_labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "id": "6e5b5db9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define tokenizer with vocab size\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "\n",
    "tokenizer = Tokenizer(num_words = vocab_size, oov_token=oov_tok)\n",
    "tokenizer.fit_on_texts(train_dataset)\n",
    "word_index = tokenizer.word_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "id": "df32e488",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6941"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Tokenize list as input\n",
    "train_sequences = tokenizer.texts_to_sequences(train_dataset)\n",
    "len(train_sequences)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "dd7422a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# truncate and pad input sequences with max_length=200\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "\n",
    "train_padded = pad_sequences(train_sequences, maxlen=max_length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "id": "d98b9981",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6941, 200)"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_padded),len(train_padded[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "id": "e0a7bf9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tokenize and pad test data\n",
    "validation_sequences = tokenizer.texts_to_sequences(test_dataset)\n",
    "validation_padded = pad_sequences(validation_sequences, maxlen=max_length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "id": "88e1ee4e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0, 1}"
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "id": "0aa06f82",
   "metadata": {},
   "outputs": [],
   "source": [
    "# re-shape train and test labels\n",
    "import numpy as np\n",
    "\n",
    "training_label_seq = np.array(train_labels)\n",
    "validation_label_seq = np.array(validation_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2198c2f1",
   "metadata": {},
   "source": [
    "LSTM"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95a97b7c",
   "metadata": {},
   "source": [
    "A bidirectional LSTM network is simply two separate LSTM networks; one feeds with a forward sequence and another with reversed sequence. Then the output of the two LSTM networks is concatenated together before being fed to the subsequent layers of the network. In Keras, you have the function Bidirectional() to clone an LSTM layer for forward-backward input and concatenate their output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "id": "99471a13",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_11\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_11 (Embedding)    (None, None, 32)          160000    \n",
      "                                                                 \n",
      " bidirectional_11 (Bidirect  (None, 64)                16640     \n",
      " ional)                                                          \n",
      "                                                                 \n",
      " dense_9 (Dense)             (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 176705 (690.25 KB)\n",
      "Trainable params: 176705 (690.25 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Define the LSTM model\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Embedding, Dropout, SimpleRNN, Dense, LSTM, Bidirectional\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(vocab_size,embedding_dim ))\n",
    "model.add(Bidirectional(LSTM(embedding_dim,dropout=0.2, recurrent_dropout=0.2)))\n",
    "#model.add(Dropout(0.5))\n",
    "model.add(Dense(1,activation='sigmoid' ))\n",
    "model.summary()\n",
    "\n",
    "# # truncate and pad input sequences\n",
    "# max_review_length = 200\n",
    "# # create the model\n",
    "# embedding_vecor_length = 32\n",
    "# model = Sequential()\n",
    "# model.add(Embedding(top_words, embedding_vecor_length, input_length=max_review_length))\n",
    "# model.add(Bidirectional(LSTM(100, dropout=0.2, recurrent_dropout=0.2)))\n",
    "# model.add(Dense(1, activation='sigmoid'))\n",
    "# model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "id": "8ed5bd90",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "import tensorflow as tf\n",
    "\n",
    "opt = tf.keras.optimizers.Adam()\n",
    "model.compile(loss='binary_crossentropy',optimizer=opt,metrics=['accuracy'])\n",
    "\n",
    "#model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "#print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "id": "fcf04986",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "217/217 - 240s - loss: 0.6631 - accuracy: 0.6195 - val_loss: 0.6600 - val_accuracy: 0.6257 - 240s/epoch - 1s/step\n",
      "Epoch 2/10\n",
      "217/217 - 227s - loss: 0.6070 - accuracy: 0.6637 - val_loss: 0.6944 - val_accuracy: 0.5938 - 227s/epoch - 1s/step\n",
      "Epoch 3/10\n",
      "217/217 - 194s - loss: 0.4971 - accuracy: 0.7531 - val_loss: 0.7804 - val_accuracy: 0.5450 - 194s/epoch - 893ms/step\n",
      "Epoch 4/10\n",
      "217/217 - 186s - loss: 0.4049 - accuracy: 0.8124 - val_loss: 0.8575 - val_accuracy: 0.5477 - 186s/epoch - 859ms/step\n",
      "Epoch 5/10\n",
      "217/217 - 192s - loss: 0.3347 - accuracy: 0.8492 - val_loss: 1.0383 - val_accuracy: 0.5598 - 192s/epoch - 883ms/step\n",
      "Epoch 6/10\n",
      "217/217 - 187s - loss: 0.2758 - accuracy: 0.8775 - val_loss: 1.1667 - val_accuracy: 0.5521 - 187s/epoch - 861ms/step\n",
      "Epoch 7/10\n",
      "217/217 - 186s - loss: 0.2412 - accuracy: 0.8911 - val_loss: 1.3370 - val_accuracy: 0.5464 - 186s/epoch - 858ms/step\n",
      "Epoch 8/10\n",
      "217/217 - 191s - loss: 0.2089 - accuracy: 0.9082 - val_loss: 1.4283 - val_accuracy: 0.5585 - 191s/epoch - 880ms/step\n",
      "Epoch 9/10\n",
      "217/217 - 193s - loss: 0.1987 - accuracy: 0.9146 - val_loss: 1.5277 - val_accuracy: 0.5403 - 193s/epoch - 889ms/step\n",
      "Epoch 10/10\n",
      "217/217 - 189s - loss: 0.1775 - accuracy: 0.9236 - val_loss: 1.5205 - val_accuracy: 0.5450 - 189s/epoch - 871ms/step\n"
     ]
    }
   ],
   "source": [
    "# Train the model\n",
    "\n",
    "num_epochs = 10\n",
    "history = model.fit(train_padded, training_label_seq, epochs=num_epochs, \n",
    "                    validation_data=(validation_padded, validation_label_seq), verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "id": "be8d0154",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plots\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "print(history.history.keys())\n",
    "#accuracy\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()\n",
    "\n",
    "#loss\n",
    "plt.plot(history.history['loss'], 'b', label = 'train')\n",
    "plt.plot(history.history['val_loss'], 'r',  label = 'test')\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epochs')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4bc1944",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}