train_mitdb.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from model_training_utils import fix_betti_string_mitdb\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tensorflow.keras.models import load_model\n",
    "import numpy as np\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix, classification_report"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Data prep"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "outputs": [],
   "source": [
    "df = pd.read_csv('data_sliced_tda/mitDB.csv')\n",
    "# type fix\n",
    "df['betti_H0'] = df['betti'].apply(lambda x: fix_betti_string_mitdb(x)[0])\n",
    "df['betti_H1'] = df['betti'].apply(lambda x: fix_betti_string_mitdb(x)[1])\n",
    "\n",
    "# Turn the list into a np.array of floats\n",
    "df['betti_H1'] = df['betti_H1'].apply(lambda x: (np.array([float(num) for num in x])))\n",
    "df['betti_H0'] = df['betti_H0'].apply(lambda x: (np.array([float(num) for num in x])))"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:40:47.161452938Z",
     "start_time": "2023-06-16T00:39:54.700967773Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [],
   "source": [
    "# Split the data into data and classes\n",
    "X, y = df['betti_H1'], df['ann']\n",
    "\n",
    "# Split to train and test data\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, stratify=y, random_state=420)\n",
    "\n",
    "# Encode the classes\n",
    "y_train = to_categorical(np.array(y_train.to_list()))\n",
    "\n",
    "# Reshape input shape\n",
    "X_train = np.array(X_train.to_list())\n",
    "X_train = X_train.reshape(X_train.shape[0], X_train.shape[1], 1)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:41:11.899815553Z",
     "start_time": "2023-06-16T00:41:11.745057840Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Model training"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "model = load_model('models_raw/4conv_1dens32.hdf5')"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "history = model.fit(X_train, y_train, batch_size=128, epochs=10, validation_split=0.2, verbose=1)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [],
   "source": [
    "model.save('model_trained/mitdb_trained.hdf5')"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:50:30.068737874Z",
     "start_time": "2023-06-16T00:50:29.941471417Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "outputs": [
    {
     "data": {
      "text/plain": "[<matplotlib.lines.Line2D at 0x7f99a0b12b30>]"
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.epoch, history.history['loss'])\n",
    "plt.plot(history.epoch, history.history['val_loss'])"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-15T22:30:23.988485322Z",
     "start_time": "2023-06-15T22:30:23.814149534Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "outputs": [
    {
     "data": {
      "text/plain": "[<matplotlib.lines.Line2D at 0x7f999cec0b80>]"
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.epoch, history.history['accuracy'])\n",
    "plt.plot(history.epoch, history.history['val_accuracy'])"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-15T22:30:24.159917098Z",
     "start_time": "2023-06-15T22:30:23.988719293Z"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Model testing and evaluating"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "outputs": [],
   "source": [
    "# Encode the test classes\n",
    "y_test = to_categorical(np.array(y_test.to_list()))"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:51:04.493335381Z",
     "start_time": "2023-06-16T00:51:04.452609279Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "683/683 [==============================] - 4s 5ms/step - loss: 0.3428 - accuracy: 0.8438\n",
      "Test Loss: 0.3428151607513428\n",
      "Test Accuracy: 0.843808650970459\n"
     ]
    }
   ],
   "source": [
    "X_test = np.array(X_test.to_list())\n",
    "X_test = X_test.reshape(X_test.shape[0], X_test.shape[1], 1)\n",
    "loss, accuracy = model.evaluate(X_test, y_test)\n",
    "print(\"Test Loss:\", loss)\n",
    "print(\"Test Accuracy:\", accuracy)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:51:09.342959384Z",
     "start_time": "2023-06-16T00:51:05.614100366Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "683/683 [==============================] - 3s 5ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": "array([[15533,  1043],\n       [ 2369,  2900]])"
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_true = np.argmax(y_test, axis=1)\n",
    "y_pred = np.argmax(model.predict(X_test), axis=1)\n",
    "\n",
    "confusion_matrix(y_true, y_pred)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:52:50.394130849Z",
     "start_time": "2023-06-16T00:52:46.887103782Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.87      0.94      0.90     16576\n",
      "           1       0.74      0.55      0.63      5269\n",
      "\n",
      "    accuracy                           0.84     21845\n",
      "   macro avg       0.80      0.74      0.77     21845\n",
      "weighted avg       0.84      0.84      0.84     21845\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_true, y_pred))"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-06-16T00:52:52.816026590Z",
     "start_time": "2023-06-16T00:52:52.805918700Z"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}