test.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Classify radar return signal from the ionosphere as __good__ and __bad__ signal. For more information, read the __ionosphere.names__ file in the foloder __data__."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from neurnet import neurnet\n",
    "from read_iono import iono_data\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "iono = iono_data()\n",
    "x_train, y_train = iono.get_train()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Design the neural network\n",
    "- __num_ins__ :number of neurons in the input layer\n",
    "\n",
    "- __num_layers__ :number of neurons in the hidden layers\n",
    "- __num_layers__ = [] if the is no hidden layer\n",
    "- __num_outs__ :number of neurons in the output layer\n",
    "- __afuns__ : list of activation functions in the network\n",
    "- __lfun__ : loss function\n",
    "    - __BCE__ for binary classification \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_ins = x_train.shape[0] \n",
    "num_layers = [10, 5]\n",
    "num_outs = y_train.shape[0]\n",
    "afuns = ['relu','relu', 'sigmoid']\n",
    "lfun = 'BCE'"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Initialize the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "nnet = neurnet(num_ins, num_layers, num_outs, x_train, afuns, y_train, lfun)\n",
    "nnet.initnet()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train the network and calculate the loss as a function of epoch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss = np.full(300, np.nan)\n",
    "for i in range(300):\n",
    "    nnet.update_network(0.25)\n",
    "    l = nnet.net_loss()[-1]\n",
    "    loss[i] = np.mean(np.abs(l[0,:]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(loss)\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Test the trained neural network on the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test, y_test = iono.get_test()\n",
    "nnet.input = x_test\n",
    "y_p = nnet.ffn()[1][-1]\n",
    "y_p[y_p<0.5] = 0\n",
    "y_p[y_p>=0.5] = 1\n",
    "y_p = y_p.astype(int)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Calculate performance of the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "90.0%\n"
     ]
    }
   ],
   "source": [
    "inds_fail = np.argwhere(np.abs(y_p[0,:]-y_test[0,:])>0)[:,0]\n",
    "size_test = y_test.size\n",
    "size_fail = inds_fail.size\n",
    "p = 100*(size_test-size_fail)/size_test\n",
    "print(str(p)+'%')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "phys",
   "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": "3.11.3"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}