autokeras.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.utils import plot_model\n",
    "import autokeras as ak\n",
    "\n",
    "import pydotplus\n",
    "import graphviz\n",
    "\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data\n",
    "url_data = 'data - 1,000.csv'\n",
    "df = pd.read_csv(url_data, index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.loc[:,'G1a':'G2b']\n",
    "y = df.loc[:, 'force']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols=['G1a','G1b','G1c','G2a','G2b']\n",
    "types=len(cols)*['numerical']\n",
    "types=dict(zip(cols, types))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "reg = ak.StructuredDataRegressor(max_trials=1, \n",
    "                                 directory='sdr',\n",
    "                                 column_names=cols, \n",
    "                                 column_types=types)\n",
    "\n",
    "reg.fit(X_train, y_train, verbose=2, epochs=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "reg.evaluate(x=X_test, y=y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = reg.export_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_model(model, show_shapes=True, expand_nested=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save architecture\n",
    "with open(\"model-arch.json\", \"w\") as file:\n",
    "    file.write(model.to_json())\n",
    "    \n",
    "# save serialized weights\n",
    "model.save_weights(\"model-weights.h5\")\n",
    "\n",
    "# save human readable weights\n",
    "with open(\"model-weights.txt\", \"w\") as file:\n",
    "    file.write(str(model.get_weights()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "config = model.to_json()\n",
    "loaded = tf.keras.models.model_from_json(config)"
   ]
  }
 ],
 "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
}