code/004_rnn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Building a RNN on SMILES data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Classifying three types of chemicals according to their MeSH therapetutic drug class.       "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<span style=\"float:left;\">\n",
    "\n",
    "|Drug Class|Number of Elements|\n",
    "|:---|:---|\n",
    "|antineoplastic|1174|\n",
    "|cns|1141|\n",
    "|cardio|787|"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this notebook, I'm running 2 types of RNN models. One Simple RNN and one LSTM RNN.\n",
    "\n",
    "- [Model 1](#model1)\n",
    "- [Model 2](#model2) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import absolute_import, division, print_function, unicode_literals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import os\n",
    "import pickle\n",
    "\n",
    "\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Embedding, SimpleRNN, Dense, Flatten, Dropout, MaxPooling2D, GRU, LSTM\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "from tensorflow.keras import regularizers\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "\n",
    "from tensorflow.keras.datasets import imdb\n",
    "from tensorflow.keras.preprocessing import sequence\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "\n",
    "%run ../assets/accuracy_loss_plots.py # function to plot accuracy and loss\n",
    "%run ../assets/confusion_matrix.py # function to plot confusion matrix\n",
    "%run ../assets/smiles_tokenizer.py # function to tokenize SMILES\n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "plt.style.use('fivethirtyeight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read in data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"../data/CID_properties_nr.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lowercase column names\n",
    "df.columns = df.columns.str.lower()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove extra classes\n",
    "index_names = df[(df[\"drug_class\"] == \"antineoplastic\") | (df[\"drug_class\"] == \"cns\") | (df[\"drug_class\"] == \"cardio\")].index\n",
    "\n",
    "df = df.loc[index_names]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "antineoplastic    1174\n",
       "cns               1141\n",
       "cardio             787\n",
       "Name: drug_class, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"drug_class\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set up X and y variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make my drug class variable categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "classes_dict = {\n",
    "    \"antineoplastic\": 0,\n",
    "    \"cardio\": 1,\n",
    "    \"cns\": 2,}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>cid</th>\n",
       "      <th>hbondacceptorcount</th>\n",
       "      <th>hbonddonorcount</th>\n",
       "      <th>isomericsmiles</th>\n",
       "      <th>molecularweight</th>\n",
       "      <th>xlogp</th>\n",
       "      <th>drug_class</th>\n",
       "      <th>desalted_smiles</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>134694070</td>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "      <td>C1CN=C(N1)NC2=C(C3=NC=CN=C3C=C2)Br.[C@@H](C(C(...</td>\n",
       "      <td>442.226</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>Brc1c(NC2=NCCN2)ccc2nccnc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4660557</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>C1C2CC3CC1CC(C2)C3NC4=CC=C(C=C4)Br</td>\n",
       "      <td>306.247</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>Brc1ccc(NC2C3CC4CC(C3)CC2C4)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>122175</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>CC(CCC(C#C)N)N</td>\n",
       "      <td>126.203</td>\n",
       "      <td>-0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>C#CC(N)CCC(C)N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5281077</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>CCC(/C=C/Cl)(C#C)O</td>\n",
       "      <td>144.598</td>\n",
       "      <td>1.5</td>\n",
       "      <td>2</td>\n",
       "      <td>C#CC(O)(/C=C/Cl)CC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>92913</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>CC(CC1=CC=CC=C1)N(C)CC#C.Cl</td>\n",
       "      <td>223.744</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>C#CCN(C)C(C)Cc1ccccc1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          cid  hbondacceptorcount  hbonddonorcount  \\\n",
       "1   134694070                   9                6   \n",
       "3     4660557                   1                1   \n",
       "4      122175                   2                2   \n",
       "5     5281077                   1                1   \n",
       "12      92913                   1                1   \n",
       "\n",
       "                                       isomericsmiles  molecularweight  xlogp  \\\n",
       "1   C1CN=C(N1)NC2=C(C3=NC=CN=C3C=C2)Br.[C@@H](C(C(...          442.226    NaN   \n",
       "3                  C1C2CC3CC1CC(C2)C3NC4=CC=C(C=C4)Br          306.247    5.0   \n",
       "4                                      CC(CCC(C#C)N)N          126.203   -0.4   \n",
       "5                                  CCC(/C=C/Cl)(C#C)O          144.598    1.5   \n",
       "12                        CC(CC1=CC=CC=C1)N(C)CC#C.Cl          223.744    NaN   \n",
       "\n",
       "    drug_class                  desalted_smiles  \n",
       "1            1      Brc1c(NC2=NCCN2)ccc2nccnc12  \n",
       "3            2  Brc1ccc(NC2C3CC4CC(C3)CC2C4)cc1  \n",
       "4            0                   C#CC(N)CCC(C)N  \n",
       "5            2               C#CC(O)(/C=C/Cl)CC  \n",
       "12           2            C#CCN(C)C(C)Cc1ccccc1  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# map drug classes to numerical\n",
    "df['drug_class'] = df['drug_class'].map(classes_dict)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1174\n",
       "2    1141\n",
       "1     787\n",
       "Name: drug_class, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check that the values are still the same\n",
    "df[\"drug_class\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set up my X and y variables\n",
    "X = df[\"desalted_smiles\"]\n",
    "y = df[\"drug_class\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split my train and test sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y,\n",
    "                                                    test_size = 0.2,\n",
    "                                                    stratify = y, random_state = 42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Check for two letter elements"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Two letter elements would be split into individual letters when going through the tokenizer. I suspect that the neural network would perform better if the tokenizer could identify the elements as one unit. Here I am checking which two letter elements appear in my SMILES."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# List of all possibble two letter elements\n",
    "two_letter_elements = ['Li','Na','Al','Si','Cl','Sc','Zn','As','Se','Br','Sn','Te','Cn']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Custom function to check for two letter elements in my list of smiles\n",
    "def check_two_letters(smile):\n",
    "    elements = []\n",
    "    for element in two_letter_elements:\n",
    "        if df[\"desalted_smiles\"].str.contains(element).sum() > 0:\n",
    "            sum = df[\"desalted_smiles\"].str.contains(element).sum()\n",
    "            elements.append(element)\n",
    "            print(f\"{element} found in {sum} rows\")\n",
    "    return elements"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Al found in 1 rows\n",
      "Cl found in 377 rows\n",
      "Sc found in 63 rows\n",
      "As found in 2 rows\n",
      "Se found in 4 rows\n",
      "Br found in 42 rows\n",
      "Cn found in 59 rows\n"
     ]
    }
   ],
   "source": [
    "# Check for which two letter elements are included in my smiles\n",
    "two_letters = check_two_letters(df[\"desalted_smiles\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Al', 'Cl', 'Sc', 'As', 'Se', 'Br', 'Cn']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "two_letters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocess Text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set up tokenizer\n",
    "tokenizer = Tokenizer(char_level=True, lower = False , oov_token=\"Cl\")\n",
    "tokenizer.fit_on_texts(df[\"desalted_smiles\"])\n",
    "sequence_of_int = tokenizer.texts_to_sequences(df[\"desalted_smiles\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tokenize text\n",
    "X_train_tk = tokenizer.texts_to_matrix(X_train, mode='count' )\n",
    "X_test_tk = tokenizer.texts_to_matrix(X_test, mode='count' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 51 unique tokens.\n"
     ]
    }
   ],
   "source": [
    "# This is how you can recover the word index that was computed\n",
    "word_index = tokenizer.word_index\n",
    "print('Found %s unique tokens.' % len(word_index))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.,  0., 82.,  0., 44., 44., 32., 27.,  2., 17., 17., 25., 17.,\n",
       "       21.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
       "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
       "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking format of np array\n",
    "X_train_tk[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Cl': 1,\n",
       " 'C': 2,\n",
       " 'c': 3,\n",
       " '(': 4,\n",
       " ')': 5,\n",
       " 'O': 6,\n",
       " '@': 7,\n",
       " '1': 8,\n",
       " '[': 9,\n",
       " ']': 10,\n",
       " '=': 11,\n",
       " 'H': 12,\n",
       " 'N': 13,\n",
       " '2': 14,\n",
       " '3': 15,\n",
       " 'n': 16,\n",
       " '4': 17,\n",
       " '/': 18,\n",
       " '5': 19,\n",
       " '-': 20,\n",
       " 'S': 21,\n",
       " 'l': 22,\n",
       " 'F': 23,\n",
       " '+': 24,\n",
       " '6': 25,\n",
       " '\\\\': 26,\n",
       " '.': 27,\n",
       " 'P': 28,\n",
       " '#': 29,\n",
       " 's': 30,\n",
       " 'o': 31,\n",
       " '7': 32,\n",
       " 'B': 33,\n",
       " 'r': 34,\n",
       " '8': 35,\n",
       " 't': 36,\n",
       " 'I': 37,\n",
       " '9': 38,\n",
       " 'e': 39,\n",
       " 'A': 40,\n",
       " 'G': 41,\n",
       " 'm': 42,\n",
       " 'M': 43,\n",
       " '%': 44,\n",
       " '0': 45,\n",
       " 'i': 46,\n",
       " 'T': 47,\n",
       " 'u': 48,\n",
       " 'a': 49,\n",
       " 'b': 50,\n",
       " 'X': 51}"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# tokenizer keeps the special characters by default! \n",
    "# This is useful for me since the special characters have significance in SMILES\n",
    "word_index"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='model1'></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model 1: Simple RNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(2020)\n",
    "model = Sequential([\n",
    "    Embedding(500,64, input_length = 52),\n",
    "    SimpleRNN(100),\n",
    "    Dense(64, activation = \"relu\"),\n",
    "    Dropout(0.5),\n",
    "    Dense(32, activation = \"relu\"),\n",
    "    Dense(3, activation = \"softmax\")\n",
    "])\n",
    "\n",
    "model.compile(optimizer = \"adam\", loss =  \"sparse_categorical_crossentropy\",\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# stop this model when it gets worse\n",
    "es = EarlyStopping(\n",
    "    monitor = \"val_loss\",\n",
    "    patience = 5, # wait 5 epochs before we really think this model is getting worse\n",
    "    min_delta = 0 # delta is the difference of validation loss from epoch to epoch. Any increase in validation loss would be considered an increase\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 2481 samples, validate on 621 samples\n",
      "Epoch 1/100\n",
      "2481/2481 - 2s - loss: 1.0747 - accuracy: 0.4184 - val_loss: 1.0341 - val_accuracy: 0.4734\n",
      "Epoch 2/100\n",
      "2481/2481 - 1s - loss: 1.0058 - accuracy: 0.5175 - val_loss: 1.0090 - val_accuracy: 0.4879\n",
      "Epoch 3/100\n",
      "2481/2481 - 1s - loss: 0.9730 - accuracy: 0.5449 - val_loss: 0.9821 - val_accuracy: 0.5153\n",
      "Epoch 4/100\n",
      "2481/2481 - 1s - loss: 0.9582 - accuracy: 0.5603 - val_loss: 0.9875 - val_accuracy: 0.5395\n",
      "Epoch 5/100\n",
      "2481/2481 - 1s - loss: 0.9211 - accuracy: 0.5804 - val_loss: 0.9784 - val_accuracy: 0.5539\n",
      "Epoch 6/100\n",
      "2481/2481 - 1s - loss: 0.9098 - accuracy: 0.5820 - val_loss: 0.9753 - val_accuracy: 0.5362\n",
      "Epoch 7/100\n",
      "2481/2481 - 1s - loss: 0.9043 - accuracy: 0.5820 - val_loss: 0.9645 - val_accuracy: 0.5556\n",
      "Epoch 8/100\n",
      "2481/2481 - 1s - loss: 0.8702 - accuracy: 0.6098 - val_loss: 0.9660 - val_accuracy: 0.5523\n",
      "Epoch 9/100\n",
      "2481/2481 - 1s - loss: 0.8672 - accuracy: 0.6098 - val_loss: 0.9461 - val_accuracy: 0.5620\n",
      "Epoch 10/100\n",
      "2481/2481 - 1s - loss: 0.8747 - accuracy: 0.5913 - val_loss: 0.9437 - val_accuracy: 0.5878\n",
      "Epoch 11/100\n",
      "2481/2481 - 1s - loss: 0.8503 - accuracy: 0.6155 - val_loss: 0.9360 - val_accuracy: 0.5572\n",
      "Epoch 12/100\n",
      "2481/2481 - 1s - loss: 0.8342 - accuracy: 0.6292 - val_loss: 0.9181 - val_accuracy: 0.6039\n",
      "Epoch 13/100\n",
      "2481/2481 - 1s - loss: 0.7979 - accuracy: 0.6485 - val_loss: 0.8996 - val_accuracy: 0.6023\n",
      "Epoch 14/100\n",
      "2481/2481 - 1s - loss: 0.8034 - accuracy: 0.6340 - val_loss: 0.9114 - val_accuracy: 0.5942\n",
      "Epoch 15/100\n",
      "2481/2481 - 1s - loss: 0.7896 - accuracy: 0.6477 - val_loss: 0.8879 - val_accuracy: 0.6103\n",
      "Epoch 16/100\n",
      "2481/2481 - 1s - loss: 0.7584 - accuracy: 0.6638 - val_loss: 0.8943 - val_accuracy: 0.6055\n",
      "Epoch 17/100\n",
      "2481/2481 - 1s - loss: 0.7468 - accuracy: 0.6606 - val_loss: 0.9059 - val_accuracy: 0.6055\n",
      "Epoch 18/100\n",
      "2481/2481 - 1s - loss: 0.7843 - accuracy: 0.6441 - val_loss: 0.9187 - val_accuracy: 0.5990\n",
      "Epoch 19/100\n",
      "2481/2481 - 1s - loss: 0.7565 - accuracy: 0.6550 - val_loss: 0.8811 - val_accuracy: 0.5958\n",
      "Epoch 20/100\n",
      "2481/2481 - 1s - loss: 0.7240 - accuracy: 0.6651 - val_loss: 0.8795 - val_accuracy: 0.6184\n",
      "Epoch 21/100\n",
      "2481/2481 - 1s - loss: 0.6896 - accuracy: 0.6965 - val_loss: 0.9343 - val_accuracy: 0.6184\n",
      "Epoch 22/100\n",
      "2481/2481 - 1s - loss: 0.6989 - accuracy: 0.6820 - val_loss: 0.8676 - val_accuracy: 0.6248\n",
      "Epoch 23/100\n",
      "2481/2481 - 1s - loss: 0.6874 - accuracy: 0.6977 - val_loss: 0.8707 - val_accuracy: 0.6409\n",
      "Epoch 24/100\n",
      "2481/2481 - 1s - loss: 0.6670 - accuracy: 0.6985 - val_loss: 0.8838 - val_accuracy: 0.6151\n",
      "Epoch 25/100\n",
      "2481/2481 - 1s - loss: 0.6937 - accuracy: 0.6860 - val_loss: 0.9334 - val_accuracy: 0.6151\n",
      "Epoch 26/100\n",
      "2481/2481 - 1s - loss: 0.7124 - accuracy: 0.6909 - val_loss: 0.8859 - val_accuracy: 0.6200\n",
      "Epoch 27/100\n",
      "2481/2481 - 1s - loss: 0.6766 - accuracy: 0.7033 - val_loss: 0.8438 - val_accuracy: 0.6473\n",
      "Epoch 28/100\n",
      "2481/2481 - 1s - loss: 0.6697 - accuracy: 0.7050 - val_loss: 0.8430 - val_accuracy: 0.6425\n",
      "Epoch 29/100\n",
      "2481/2481 - 1s - loss: 0.6387 - accuracy: 0.7231 - val_loss: 0.8667 - val_accuracy: 0.6361\n",
      "Epoch 30/100\n",
      "2481/2481 - 1s - loss: 0.6113 - accuracy: 0.7255 - val_loss: 0.8831 - val_accuracy: 0.6425\n",
      "Epoch 31/100\n",
      "2481/2481 - 1s - loss: 0.6039 - accuracy: 0.7368 - val_loss: 0.8455 - val_accuracy: 0.6473\n",
      "Epoch 32/100\n",
      "2481/2481 - 1s - loss: 0.5843 - accuracy: 0.7428 - val_loss: 0.9474 - val_accuracy: 0.6441\n",
      "Epoch 33/100\n",
      "2481/2481 - 1s - loss: 0.5793 - accuracy: 0.7441 - val_loss: 0.9096 - val_accuracy: 0.6425\n"
     ]
    }
   ],
   "source": [
    "# fit model\n",
    "history = model.fit(X_train_tk, y_train,\n",
    "                   epochs = 100,\n",
    "                   batch_size = 128, \n",
    "                   validation_data = (X_test_tk, y_test),\n",
    "                   verbose = 2,\n",
    "                   callbacks = [es]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 479,
       "width": 979
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot accuracy and loss for model\n",
    "plot_accuracy_loss(history, epochs = 33);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Generate predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions on test set\n",
    "preds = model.predict_classes(X_test_tk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "val_acc = round(history.history['val_accuracy'][-1], 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[171  17  47]\n",
      " [ 39  76  43]\n",
      " [ 37  39 152]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 496,
       "width": 492
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Look at confusion matrix\n",
    "cnf_matrix = confusion_matrix(y_test, preds,labels=[0, 1, 2])\n",
    "np.set_printoptions(precision=2)\n",
    "\n",
    "# Plot non-normalized confusion matrix\n",
    "plt.figure(figsize = (7, 7))\n",
    "plot_confusion_matrix(cnf_matrix, classes=['Antineoplastic','Cardio',  'CNS'],\n",
    "                      title=f'Confusion Matrix for RNN Model \\nAccuracy Score: {val_acc}\\n' )\n",
    "plt.savefig(\"../plots/confusion_matrix_rnn.png\", dpi = 150);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [],
   "source": [
    "#save model\n",
    "model3.save(\"../assets/model_rnn.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='model2'></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 2: LSTM Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Comparing how a LSTM model performs to simple RNN. My hypothesis is that LSTM model would perform better because it's supposed to hold more short term memory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 2481 samples, validate on 621 samples\n",
      "Epoch 1/20\n",
      "2481/2481 - 31s - loss: 1.0892 - acc: 0.3583 - val_loss: 1.0855 - val_acc: 0.3784\n",
      "Epoch 2/20\n",
      "2481/2481 - 24s - loss: 1.0858 - acc: 0.3680 - val_loss: 1.0841 - val_acc: 0.3784\n",
      "Epoch 3/20\n",
      "2481/2481 - 21s - loss: 1.0848 - acc: 0.3531 - val_loss: 1.0839 - val_acc: 0.3784\n",
      "Epoch 4/20\n",
      "2481/2481 - 19s - loss: 1.0840 - acc: 0.3785 - val_loss: 1.0839 - val_acc: 0.3784\n",
      "Epoch 5/20\n",
      "2481/2481 - 24s - loss: 1.0845 - acc: 0.3692 - val_loss: 1.0841 - val_acc: 0.3784\n",
      "Epoch 6/20\n",
      "2481/2481 - 22s - loss: 1.0841 - acc: 0.3785 - val_loss: 1.0850 - val_acc: 0.3784\n",
      "Epoch 7/20\n",
      "2481/2481 - 21s - loss: 1.0855 - acc: 0.3668 - val_loss: 1.0839 - val_acc: 0.3784\n",
      "Epoch 8/20\n",
      "2481/2481 - 15s - loss: 1.0839 - acc: 0.3769 - val_loss: 1.0843 - val_acc: 0.3671\n",
      "Epoch 9/20\n",
      "2481/2481 - 16s - loss: 1.0839 - acc: 0.3696 - val_loss: 1.0838 - val_acc: 0.3784\n",
      "Epoch 10/20\n",
      "2481/2481 - 29s - loss: 1.0839 - acc: 0.3785 - val_loss: 1.0841 - val_acc: 0.3784\n",
      "Epoch 11/20\n",
      "2481/2481 - 26s - loss: 1.1014 - acc: 0.3821 - val_loss: 1.0839 - val_acc: 0.3784\n",
      "Epoch 12/20\n",
      "2481/2481 - 16s - loss: 1.0841 - acc: 0.3744 - val_loss: 1.0841 - val_acc: 0.3784\n",
      "Epoch 13/20\n",
      "2481/2481 - 14s - loss: 1.0839 - acc: 0.3785 - val_loss: 1.0841 - val_acc: 0.3784\n",
      "Epoch 14/20\n",
      "2481/2481 - 13s - loss: 1.0843 - acc: 0.3785 - val_loss: 1.0839 - val_acc: 0.3784\n",
      "Epoch 15/20\n",
      "2481/2481 - 14s - loss: 1.0838 - acc: 0.3785 - val_loss: 1.0838 - val_acc: 0.3784\n",
      "Epoch 16/20\n",
      "2481/2481 - 14s - loss: 1.0839 - acc: 0.3676 - val_loss: 1.0838 - val_acc: 0.3784\n",
      "Epoch 17/20\n",
      "2481/2481 - 15s - loss: 1.0837 - acc: 0.3789 - val_loss: 1.0838 - val_acc: 0.3784\n",
      "Epoch 18/20\n",
      "2481/2481 - 14s - loss: 1.0840 - acc: 0.3789 - val_loss: 1.0839 - val_acc: 0.3784\n",
      "Epoch 19/20\n",
      "2481/2481 - 14s - loss: 1.0841 - acc: 0.3789 - val_loss: 1.0838 - val_acc: 0.3784\n",
      "Epoch 20/20\n",
      "2481/2481 - 14s - loss: 1.0839 - acc: 0.3789 - val_loss: 1.0839 - val_acc: 0.3784\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Embedding(500, 64, input_length = 52))\n",
    "model.add(LSTM(512))\n",
    "Dropout(0.5)\n",
    "model.add(Dense(3, activation='softmax'))\n",
    "\n",
    "model.compile(optimizer='adam',\n",
    "              loss=\"sparse_categorical_crossentropy\",\n",
    "              metrics=['acc'])\n",
    "\n",
    "history = model.fit(X_train_tk, y_train,\n",
    "                   epochs = 20,\n",
    "                   batch_size = 128, \n",
    "                   validation_data = (X_test_tk, y_test),\n",
    "                   verbose = 2,)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I could fine tune this model more, but it appears that it is performing much worse than the simple RNN model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Model 1, the one using simple RNN has the highest accuracy score, without too much overfitting."
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}