poc/nnlm_tfidf.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# POC: Using tfidf and NNLM and word2vec for SDTM mapping\n",
    "Sam Tomioka\n",
    "\n",
    "Note: This is very rough notebook.\n",
    "\n",
    "Data created with R.\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow_hub as hub\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.model_selection import cross_val_predict\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import cohen_kappa_score\n",
    "from sklearn.metrics import precision_recall_fscore_support\n",
    "from sklearn.metrics import classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"train_data/ae_meta.csv\")\n",
    "df.head\n",
    "X = df[['ID', 'text']]\n",
    "y = df['sdtm'].tolist()\n",
    "\n",
    "(X_train, X_test, y_train, y_test) = train_test_split(\n",
    "    X, y, test_size=0.2, random_state=2018,\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TFIDF alone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/model_selection/_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.6405821627137254"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe = Pipeline([('tfidf',TfidfVectorizer(min_df=3, max_df=10)), \n",
    "                 (\"clf\", LogisticRegression())])\n",
    "np.mean(cross_val_score(pipe, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/model_selection/_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.6025520108264882"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe = Pipeline([('tfidf',TfidfVectorizer(min_df=3, max_df=10)), (\"clf\", AdaBoostClassifier( learning_rate=0.01, n_estimators = 500))])\n",
    "np.mean(cross_val_score(pipe, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/model_selection/_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/model_selection/_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.7692993387444849, 0.6846815626817213)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe = Pipeline([('tfidf',TfidfVectorizer(min_df=3, max_df=10)), (\"clf\", GradientBoostingClassifier(n_estimators= 500, learning_rate=0.01))])\n",
    "np.mean(cross_val_score(pipe, X_train[\"ID\"], y_train, cv=10)), np.mean(cross_val_score(pipe, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NNLM + Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Using /tmp/tfhub_modules to cache modules.\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/model_selection/_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.7924854841580011"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def texts_encoder(texts):\n",
    "    with tf.Graph().as_default():\n",
    "        embed = hub.Module(\"https://tfhub.dev/google/nnlm-en-dim128-with-normalization/1\")\n",
    "        embeddings = embed(texts)\n",
    "        with tf.Session() as sess:\n",
    "            sess.run(tf.global_variables_initializer())\n",
    "            sess.run(tf.tables_initializer())\n",
    "            result = sess.run(embeddings)\n",
    "    return result\n",
    "\n",
    "\n",
    "class  MeanEmbeddingVectorizer:\n",
    "    def __init__(self):\n",
    "        pass\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        if isinstance(X, pd.core.series.Series):\n",
    "            X = X.tolist()\n",
    "        return texts_encoder(X)\n",
    "\n",
    "pipe = Pipeline([\n",
    "    ('w2v',MeanEmbeddingVectorizer()), \n",
    "    (\"clf\", LogisticRegression())])\n",
    "\n",
    "np.mean(cross_val_score(pipe, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.7150997150997151"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe.fit(X_train[\"text\"], y_train)\n",
    "# predict test instances\n",
    "y_preds = pipe.predict(X_test[\"text\"])\n",
    "accuracy_score(y_test, y_preds, normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Micro precision:  0.7150997150997151 recall:  0.7150997150997151 F-beta:  0.7150997150997151\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "      AEACN       1.00      0.67      0.80         6\n",
      "   AEACNDEV       0.00      0.00      0.00         1\n",
      "   AEACNOTH       0.00      0.00      0.00         2\n",
      "      AECAT       0.00      0.00      0.00         1\n",
      "   AECONTRT       1.00      0.50      0.67         2\n",
      "    AEDECOD       1.00      0.17      0.29         6\n",
      "      AEDIR       0.50      1.00      0.67         1\n",
      "    AEENDTC       0.00      0.00      0.00         8\n",
      " AEENDTC_TM       0.00      0.00      0.00         2\n",
      "     AEENDY       0.00      0.00      0.00         1\n",
      "     AEENRF       0.00      0.00      0.00         1\n",
      "   AEENRTPT       1.00      1.00      1.00         6\n",
      "     AEHLGT       1.00      0.14      0.25         7\n",
      "   AEHLGTCD       0.00      0.00      0.00         6\n",
      "      AEHLT       0.50      1.00      0.67         1\n",
      "    AEHLTCD       0.00      0.00      0.00         2\n",
      "      AELAT       0.00      0.00      0.00         4\n",
      "    AELIFTH       0.00      0.00      0.00         0\n",
      "      AELLT       0.33      1.00      0.50         1\n",
      "    AELLTCD       0.75      0.75      0.75         4\n",
      "   AEMODIFY       0.00      0.00      0.00         1\n",
      "      AEOUT       0.75      1.00      0.86         3\n",
      "     AEPATT       0.33      0.50      0.40         2\n",
      "     AEPTCD       0.60      0.43      0.50         7\n",
      "      AEREL       1.00      0.67      0.80         3\n",
      "    AESCONG       0.00      0.00      0.00         6\n",
      "   AESDISAB       1.00      0.33      0.50         3\n",
      "      AESER       0.00      0.00      0.00         7\n",
      "      AESEV       0.50      0.33      0.40         3\n",
      "    AESHOSP       0.67      0.50      0.57         4\n",
      "    AESLIFE       0.00      0.00      0.00         1\n",
      "     AESMIE       1.00      1.00      1.00         2\n",
      "      AESOC       0.50      0.20      0.29         5\n",
      "    AESOCCD       0.75      0.50      0.60         6\n",
      "    AESTDTC       0.00      0.00      0.00         6\n",
      "     AESTRF       0.00      0.00      0.00         2\n",
      "     AETERM       1.00      0.75      0.86         8\n",
      "    AETOXGR       0.00      0.00      0.00         1\n",
      "       DROP       0.70      1.00      0.82       193\n",
      "     DTHDTC       0.00      0.00      0.00         2\n",
      "  QNAM_AESI       0.00      0.00      0.00         1\n",
      "  QNAM_COPD       0.00      0.00      0.00         1\n",
      " QNAM_ESAM1       0.00      0.00      0.00         1\n",
      " QNAM_ESAM2       1.00      0.50      0.67         2\n",
      " QNAM_ESAM3       0.00      0.00      0.00         1\n",
      "  QNAM_EXER       0.00      0.00      0.00         1\n",
      " QNAM_EXLAB       0.00      0.00      0.00         1\n",
      "QNAM_EXSTER       0.00      0.00      0.00         1\n",
      "     SITEID       1.00      0.75      0.86         4\n",
      "    STUDYID       1.00      1.00      1.00         4\n",
      "     SUBJID       1.00      1.00      1.00         6\n",
      "SUPPAE.QVAL       0.00      0.00      0.00         1\n",
      "\n",
      "avg / total       0.62      0.72      0.63       351\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/metrics/classification.py:1137: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples.\n",
      "  'recall', 'true', average, warn_for)\n"
     ]
    }
   ],
   "source": [
    "metrics=precision_recall_fscore_support(y_test, y_preds, average='micro')\n",
    "print(\"Micro precision: \",metrics[0],\"recall: \",metrics[1],\"F-beta: \",metrics[2])\n",
    "print(classification_report(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# NNLM+Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/model_selection/_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.8939606053786553"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def texts_encoder(texts):\n",
    "    with tf.Graph().as_default():\n",
    "        embed = hub.Module(\"https://tfhub.dev/google/nnlm-en-dim128-with-normalization/1\")\n",
    "        embeddings = embed(texts)\n",
    "        with tf.Session() as sess:\n",
    "            sess.run(tf.global_variables_initializer())\n",
    "            sess.run(tf.tables_initializer())\n",
    "            result = sess.run(embeddings)\n",
    "    return result\n",
    "\n",
    "\n",
    "class  MeanEmbeddingVectorizer:\n",
    "    def __init__(self):\n",
    "        pass\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        if isinstance(X, pd.core.series.Series):\n",
    "            X = X.tolist()\n",
    "        return texts_encoder(X)\n",
    "\n",
    "pipe_gb = Pipeline([\n",
    "    ('w2v',MeanEmbeddingVectorizer()), \n",
    "    (\"clf\", GradientBoostingClassifier(loss='deviance', \n",
    "                                       learning_rate=0.01, \n",
    "                                       n_estimators=500, \n",
    "                                       subsample=0.8, \n",
    "                                       criterion='friedman_mse', \n",
    "                                       min_samples_split=2, \n",
    "                                       min_samples_leaf=1, \n",
    "                                       min_weight_fraction_leaf=0.0, \n",
    "                                       max_depth=3, \n",
    "                                       min_impurity_decrease=0.0, \n",
    "                                       verbose=0, \n",
    "                                       max_leaf_nodes=None, warm_start=False, presort='auto'))])\n",
    "np.mean(cross_val_score(pipe_gb, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "pipe_gb.fit(X_train[\"text\"], y_train)\n",
    "# predict test instances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.8774928774928775"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_preds = pipe_gb.predict(X_test[\"text\"])\n",
    "accuracy_score(y_test, y_preds, normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Micro precision:  0.8774928774928775 recall:  0.8774928774928775 F-beta:  0.8774928774928775\n"
     ]
    }
   ],
   "source": [
    "metrics=precision_recall_fscore_support(y_test, y_preds, average='micro')\n",
    "print(\"Micro precision: \",metrics[0],\"recall: \",metrics[1],\"F-beta: \",metrics[2])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "      AEACN       1.00      0.83      0.91         6\n",
      "   AEACNDEV       1.00      1.00      1.00         1\n",
      "   AEACNOTH       0.00      0.00      0.00         2\n",
      "      AECAT       0.00      0.00      0.00         1\n",
      "   AECONTRT       1.00      0.50      0.67         2\n",
      "    AEDECOD       1.00      0.33      0.50         6\n",
      "      AEDIR       1.00      1.00      1.00         1\n",
      "    AEENDTC       1.00      0.50      0.67         8\n",
      " AEENDTC_TM       0.50      1.00      0.67         2\n",
      "     AEENDY       1.00      1.00      1.00         1\n",
      "     AEENRF       0.00      0.00      0.00         1\n",
      "   AEENRTPT       0.86      1.00      0.92         6\n",
      "    AEENTPT       0.00      0.00      0.00         0\n",
      "     AEHLGT       1.00      0.71      0.83         7\n",
      "   AEHLGTCD       0.62      0.83      0.71         6\n",
      "      AEHLT       1.00      1.00      1.00         1\n",
      "    AEHLTCD       0.00      0.00      0.00         2\n",
      "      AELAT       1.00      0.75      0.86         4\n",
      "      AELLT       0.50      1.00      0.67         1\n",
      "    AELLTCD       0.67      0.50      0.57         4\n",
      "   AEMODIFY       0.00      0.00      0.00         1\n",
      "      AEOUT       1.00      1.00      1.00         3\n",
      "     AEPATT       0.67      1.00      0.80         2\n",
      "     AEPTCD       1.00      1.00      1.00         7\n",
      "      AEREL       1.00      1.00      1.00         3\n",
      "   AERELNST       0.00      0.00      0.00         0\n",
      "     AESCAT       0.00      0.00      0.00         0\n",
      "    AESCONG       1.00      1.00      1.00         6\n",
      "   AESDISAB       1.00      0.67      0.80         3\n",
      "      AESER       1.00      1.00      1.00         7\n",
      "      AESEV       1.00      0.67      0.80         3\n",
      "    AESHOSP       1.00      1.00      1.00         4\n",
      "    AESLIFE       1.00      1.00      1.00         1\n",
      "     AESMIE       1.00      0.50      0.67         2\n",
      "      AESOC       1.00      0.80      0.89         5\n",
      "    AESOCCD       1.00      0.67      0.80         6\n",
      "      AESOD       0.00      0.00      0.00         0\n",
      "    AESTDTC       0.71      0.83      0.77         6\n",
      "     AESTRF       1.00      0.50      0.67         2\n",
      "     AETERM       1.00      0.88      0.93         8\n",
      "    AETOXGR       0.00      0.00      0.00         1\n",
      "       DROP       0.89      0.99      0.94       193\n",
      "     DTHDTC       1.00      1.00      1.00         2\n",
      "  QNAM_AESI       0.00      0.00      0.00         1\n",
      "  QNAM_COPD       0.00      0.00      0.00         1\n",
      " QNAM_ESAM1       1.00      1.00      1.00         1\n",
      " QNAM_ESAM2       0.50      0.50      0.50         2\n",
      " QNAM_ESAM3       0.00      0.00      0.00         1\n",
      "  QNAM_EXER       0.00      0.00      0.00         1\n",
      " QNAM_EXLAB       1.00      1.00      1.00         1\n",
      "QNAM_EXSTER       0.00      0.00      0.00         1\n",
      "     SITEID       1.00      0.75      0.86         4\n",
      "    STUDYID       1.00      1.00      1.00         4\n",
      "     SUBJID       1.00      1.00      1.00         6\n",
      "SUPPAE.QVAL       0.00      0.00      0.00         1\n",
      "\n",
      "avg / total       0.87      0.88      0.86       351\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/home/ec2-user/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/sklearn/metrics/classification.py:1137: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples.\n",
      "  'recall', 'true', average, warn_for)\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x648 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,9))\n",
    "ax.scatter(y_test, y_preds, edgecolors=(0,0,0), alpha=0.5, color='green')\n",
    "#ax.plot([y.min(), y.max()],[y.min(),y.max()], 'k--', lw=4)\n",
    "ax.set_xlabel('Correct Mapping', )\n",
    "plt.xticks(rotation=90)\n",
    "ax.set_ylabel('Predicted')\n",
    "plt.title('NNLM+Gradient Boosting')\n",
    "\n",
    "plt.setp(ax.spines.values(),alpha=0.5, color='green' )\n",
    "plt.setp([ax.get_xticklines(), ax.get_yticklines()],alpha=0.5, color='green')\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"images/nnlm_gradient.png\", dpi=150)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TFIDF  + Logistic Regression\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Sam\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.8190993871461456"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "class  TfIdfNNLMVectorizer:\n",
    "    def __init__(self):\n",
    "        pass\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        self.tfidf_model = TfidfVectorizer(min_df=3, max_df=10).fit(X)\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        return self.tfidf_and_w2v(X, self.tfidf_model)\n",
    "\n",
    "    def texts_encoder(self, texts):\n",
    "        with tf.Graph().as_default():\n",
    "            embed = hub.Module(\"https://tfhub.dev/google/nnlm-en-dim128-with-normalization/1\")\n",
    "            embeddings = embed(texts)\n",
    "            with tf.Session() as sess:\n",
    "                sess.run(tf.global_variables_initializer())\n",
    "                sess.run(tf.tables_initializer())\n",
    "                result = sess.run(embeddings)\n",
    "        return result\n",
    "\n",
    "    def tfidf_and_w2v(self, X, tfidf_model):\n",
    "        tmp = tfidf_model.transform(X)\n",
    "        a = tmp.toarray()\n",
    "        if isinstance(X, pd.core.series.Series):\n",
    "            X = X.tolist()\n",
    "        b = self.texts_encoder(X)\n",
    "        return np.hstack((a, b))\n",
    "\n",
    "pipe_1 = Pipeline([\n",
    "    ('vector',TfIdfNNLMVectorizer()), \n",
    "     (\"clf\", LogisticRegression())])\n",
    "np.mean(cross_val_score(pipe_1, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Sam\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    }
   ],
   "source": [
    "predicted =  cross_val_predict(pipe_1, X_train[\"text\"], y_train, cv=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1b7d74512e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,9))\n",
    "ax.scatter(y_train, predicted, edgecolors=(0,0,0), alpha=0.5, color='green')\n",
    "#ax.plot([y.min(), y.max()],[y.min(),y.max()], 'k--', lw=4)\n",
    "ax.set_xlabel('Correct Mapping', )\n",
    "plt.xticks(rotation=90)\n",
    "ax.set_ylabel('Predicted')\n",
    "plt.title('Ensemble ModelA + ModelB')\n",
    "\n",
    "plt.setp(ax.spines.values(),alpha=0.5, color='green' )\n",
    "plt.setp([ax.get_xticklines(), ax.get_yticklines()],alpha=0.5, color='green')\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"tfidf_nnlm.png\", dpi=150)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Sam\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.9142281679608132"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "class  TfIdfNNLMVectorizer:\n",
    "    def __init__(self):\n",
    "        pass\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        self.tfidf_model = TfidfVectorizer(min_df=3, max_df=10).fit(X)\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        return self.tfidf_and_nnlm(X, self.tfidf_model)\n",
    "\n",
    "    def texts_encoder(self, texts):\n",
    "        with tf.Graph().as_default():\n",
    "            embed = hub.Module(\"https://tfhub.dev/google/nnlm-en-dim128-with-normalization/1\")\n",
    "            embeddings = embed(texts)\n",
    "            with tf.Session() as sess:\n",
    "                sess.run(tf.global_variables_initializer())\n",
    "                sess.run(tf.tables_initializer())\n",
    "                result = sess.run(embeddings)\n",
    "        return result\n",
    "\n",
    "    def TfidfVectorizer(self, X, tfidf_model):\n",
    "        tmp = tfidf_model.transform(X)\n",
    "        a = tmp.toarray()\n",
    "        if isinstance(X, pd.core.series.Series):\n",
    "            X = X.tolist()\n",
    "        b = self.texts_encoder(X)\n",
    "        return np.hstack((a, b))\n",
    "\n",
    "pipe3 = Pipeline([\n",
    "    ('vector',TfIdfNNLMVectorizer()), \n",
    "    (\"clf\", GradientBoostingClassifier(loss='deviance', \n",
    "                                       learning_rate=0.01, \n",
    "                                       n_estimators=500, \n",
    "                                       subsample=0.8, \n",
    "                                       criterion='friedman_mse', \n",
    "                                       min_samples_split=2, \n",
    "                                       min_samples_leaf=1, \n",
    "                                       min_weight_fraction_leaf=0.0, \n",
    "                                       max_depth=3, \n",
    "                                       min_impurity_decrease=0.0, \n",
    "                                       verbose=0, \n",
    "                                       max_leaf_nodes=None, warm_start=False, presort='auto'))])\n",
    "np.mean(cross_val_score(pipe3, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Sam\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.6237311980481719"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "class  TfIdfNNLMVectorizer:\n",
    "    def __init__(self):\n",
    "        pass\n",
    "\n",
    "    def fit(self, X, y):\n",
    "        self.tfidf_model = TfidfVectorizer(min_df=3, max_df=10).fit(X)\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        return self.TfidfVectorizer(X, self.tfidf_model)\n",
    "\n",
    "    def texts_encoder(self, texts):\n",
    "        with tf.Graph().as_default():\n",
    "            embed = hub.Module(\"https://tfhub.dev/google/nnlm-en-dim128-with-normalization/1\")\n",
    "            embeddings = embed(texts)\n",
    "            with tf.Session() as sess:\n",
    "                sess.run(tf.global_variables_initializer())\n",
    "                sess.run(tf.tables_initializer())\n",
    "                result = sess.run(embeddings)\n",
    "        return result\n",
    "\n",
    "    def TfidfVectorizer(self, X, tfidf_model):\n",
    "        tmp = tfidf_model.transform(X)\n",
    "        a = tmp.toarray()\n",
    "        if isinstance(X, pd.core.series.Series):\n",
    "            X = X.tolist()\n",
    "        b = self.texts_encoder(X)\n",
    "        return np.hstack((a, b))\n",
    "\n",
    "pipe4 = Pipeline([\n",
    "    ('vector',TfIdfNNLMVectorizer()), \n",
    "    (\"clf\", AdaBoostClassifier(learning_rate=0.01))])\n",
    "np.mean(cross_val_score(pipe4, X_train[\"text\"], y_train, cv=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Sam\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=10.\n",
      "  % (min_groups, self.n_splits)), Warning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n",
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    }
   ],
   "source": [
    "predicted =  cross_val_predict(pipe, X_train[\"text\"], y_train, cv=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('vector', <__main__.TfIdfWord2VecVectorizer object at 0x000001B6287FB0F0>), ('clf', GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.01, loss='deviance', max_depth=3,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity...      presort='auto', random_state=None, subsample=0.8, verbose=0,\n",
       "              warm_start=False))])"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe3.fit(X_train[\"text\"], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.886039886039886"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# predict test instances\n",
    "y_preds = pipe3.predict(X_test[\"text\"])\n",
    "accuracy_score(y_test, y_preds, normalize=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['DROP', 'DROP', 'DROP', 'AESTDTC', 'AEHLGT', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'DROP', 'AEENDTC', 'AEACNOTH', 'DROP', 'AEHLGTCD',\n",
       "       'SUBJID', 'AESOC', 'DROP', 'DROP', 'DROP', 'STUDYID', 'AEACN',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'AEHLGTCD', 'DROP', 'QNAM_ESAM2',\n",
       "       'DROP', 'AEHLGTCD', 'AEACNDEV', 'DROP', 'DROP', 'DROP', 'AELAT',\n",
       "       'AETERM', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP', 'AEPTCD',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'AELAT', 'AEPTCD', 'DROP',\n",
       "       'AESTRF', 'AELAT', 'DROP', 'DROP', 'QNAM_AESI', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'DROP', 'AEHLGT', 'AETERM', 'DROP', 'DROP',\n",
       "       'SITEID', 'AEPTCD', 'QNAM_ESAM1', 'DROP', 'DROP', 'DROP',\n",
       "       'AEHLGTCD', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'AEENRTPT; AEENRF', 'AESTDTC', 'STUDYID', 'DROP', 'DROP', 'AESOC',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'QNAM_COPD', 'DROP', 'AESER',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'AEPTCD', 'AEHLT', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'DROP', 'AEENRTPT', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'AESDISAB', 'DTHDTC', 'DROP', 'AESCONG', 'DROP', 'AEPATT',\n",
       "       'AEENRTPT', 'DROP', 'QNAM_EXSTER', 'AELLTCD', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'STUDYID', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'STUDYID', 'AEHLGTCD', 'AEPTCD', 'DROP', 'AEHLGT', 'DROP',\n",
       "       'AEENDTC', 'DROP', 'DROP', 'AESTDTC', 'DROP', 'AEENDY', 'AESHOSP',\n",
       "       'AEHLGTCD', 'AETERM', 'QNAM_COPD', 'AESMIE', 'AESDISAB', 'DROP',\n",
       "       'DROP', 'AEENRTPT', 'DROP', 'DROP', 'DROP', 'SUBJID', 'DROP',\n",
       "       'DROP', 'AESTDTC', 'AELLTCD', 'DROP', 'AESOCCD', 'AEREL',\n",
       "       'AELLTCD', 'AELLTCD', 'DROP', 'AEOUT', 'DROP', 'DROP', 'DROP',\n",
       "       'AEENRTPT', 'SUBJID', 'AESER', 'DROP', 'AESOC', 'DROP', 'AEDECOD',\n",
       "       'DROP', 'AEHLTCD', 'DROP', 'AESER', 'AESER', 'AEACN', 'DROP',\n",
       "       'DROP', 'AESCONG', 'DROP', 'DROP', 'DROP', 'AESCONG', 'AESOCCD',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'AEENRTPT', 'DROP', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'AEPATT', 'DROP', 'DROP', 'DROP', 'DROP', 'AEOUT',\n",
       "       'DROP', 'DROP', 'DROP', 'AEREL', 'AESEV', 'DROP', 'DROP', 'AELAT',\n",
       "       'STUDYID', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP', 'AESOC',\n",
       "       'AEACN', 'AESHOSP', 'AETERM', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'DROP', 'AEENDTC_TM', 'DROP', 'AEHLGT', 'DROP', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'DROP', 'AEPTCD', 'AESHOSP', 'AESLIFE', 'DROP',\n",
       "       'AELLT', 'AESTDTC', 'DROP', 'DROP', 'AESOCCD', 'AESER', 'AEACN',\n",
       "       'AETERM', 'AESTDTC', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'AESCONG', 'AESOCCD', 'AEENRTPT', 'DROP', 'DROP', 'DROP',\n",
       "       'AESCONG', 'SUBJID', 'DROP', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'AEDIR', 'AEHLGT', 'AEOUT', 'DROP', 'AEACN', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'QNAM_EXLAB', 'AECONTRT', 'DROP', 'DROP',\n",
       "       'AEENDTC_TM', 'DROP', 'DROP', 'AEENRTPT', 'DROP', 'DROP', 'AESEV',\n",
       "       'DROP', 'AEHLT', 'DROP', 'AETERM', 'AEPTCD', 'DROP', 'DROP',\n",
       "       'DROP', 'DROP', 'AELLT', 'DROP', 'AEENDTC', 'DROP', 'AEENDTC',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'AESHOSP', 'DROP', 'SUBJID',\n",
       "       'DROP', 'AEPATT', 'DROP', 'DROP', 'DROP', 'AESTDTC', 'DROP',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP', 'SUBJID', 'DROP', 'AETERM', 'DROP',\n",
       "       'SITEID', 'AEHLT', 'DTHDTC', 'DROP', 'AEDECOD', 'DROP', 'AESCONG',\n",
       "       'DROP', 'AESER', 'DROP', 'AEREL', 'DROP', 'DROP', 'DROP', 'DROP',\n",
       "       'SITEID', 'AESER', 'AEACNDEV', 'DROP', 'DROP', 'QNAM_ESAM1',\n",
       "       'DROP', 'DROP', 'DROP', 'DROP'], dtype='<U16')"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8276031434184676"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohen_kappa_score(y_preds, y_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1b7d4a77208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,9))\n",
    "ax.scatter(y, predicted, edgecolors=(0,0,0), alpha=0.5, color='green')\n",
    "#ax.plot([y.min(), y.max()],[y.min(),y.max()], 'k--', lw=4)\n",
    "ax.set_xlabel('Correct Mapping', )\n",
    "plt.xticks(rotation=90)\n",
    "ax.set_ylabel('Predicted')\n",
    "plt.title('Ensemble ModelA + ModelB')\n",
    "\n",
    "plt.setp(ax.spines.values(),alpha=0.5, color='green' )\n",
    "plt.setp([ax.get_xticklines(), ax.get_yticklines()],alpha=0.5, color='green')\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"tfidf_word2vec.png\", dpi=150)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1b7d1000be0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,9))\n",
    "ax.scatter(y_test, y_preds, edgecolors=(0,0,0), alpha=0.5, color='green')\n",
    "#ax.plot([y.min(), y.max()],[y.min(),y.max()], 'k--', lw=4)\n",
    "ax.set_xlabel('Correct Mapping', )\n",
    "plt.xticks(rotation=90)\n",
    "ax.set_ylabel('Predicted')\n",
    "plt.title('Ensemble ModelA + ModelB on Test Data')\n",
    "\n",
    "plt.setp(ax.spines.values(),alpha=0.5, color='green' )\n",
    "plt.setp([ax.get_xticklines(), ax.get_yticklines()],alpha=0.5, color='green')\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"tfidf_word2vec-test.png\", dpi=150)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "conda_tensorflow_p36",
   "language": "python",
   "name": "conda_tensorflow_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}