final_analysis_cs230_formatting_confmat_sensspec.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pickle\n",
    "\n",
    "def save_params(parameters, file):\n",
    "    with open(file, 'wb') as f:\n",
    "        pickle.dump(parameters, f)\n",
    "\n",
    "def load_params(file):\n",
    "    with open(file, 'rb') as f:\n",
    "        parameters = pickle.load(f)\n",
    "    return parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAADoCAYAAAD1wHPIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1cb50278358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "params = load_params('specAndsens_final.pkl')\n",
    "sens = params[0]\n",
    "specs = params[1]\n",
    "\n",
    "\n",
    "plt.figure(num=None, figsize=(5, 3), dpi=80, facecolor='w', edgecolor='k')\n",
    "plt.plot(sens,specs)\n",
    "plt.xlabel('Sensitivity')\n",
    "plt.ylabel('Specificity')\n",
    "plt.tight_layout()\n",
    "plt.savefig('sens_spec_figure.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "average_conf_mat = [[.805, .02333],[0.1,.071666]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000,)\n",
      "[ 0.00774362 -0.01109694  0.01098853 -0.00132516 -0.00260519  0.00702594\n",
      " -0.0054315  -0.0027897   0.00680439  0.00279195]\n",
      "(511,)\n",
      "0.000273793\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAACsCAYAAAAuVDhiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nS1DGGGOCkiUoY4wxQen/A6lVZWIkVgp9AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x29e79441fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = load_params(\"best_X_params.pkl\")\n",
    "X_mag = np.mean(X, axis = 2)\n",
    "X_mag = np.mean(np.mean(X_mag, axis = 2), axis = 0)\n",
    "print(X_mag.shape)\n",
    "print(X_mag[0:10])\n",
    "X_fft = np.real(np.fft.fft(X_mag, n = 1024))\n",
    "print(X_fft[1:512].shape)\n",
    "X_fft_back = np.flip(X_fft[1:512],axis = 0)\n",
    "X_fft_pos = X_fft_back+X_fft[513:1024]\n",
    "X_fft = np.insert(X_fft_pos,0,X_fft[512])\n",
    "print(np.mean(X_mag))\n",
    "\n",
    "plt.close('all')\n",
    "plt.figure(figsize=(6,2.5))\n",
    "plt.plot(np.abs(X_fft[0:10]))\n",
    "plt.ylabel(\"Fourier Component (|$A$|)\")\n",
    "plt.xlabel(\"Frequency (f)\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"cs230_fourier_generate_X.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2 3 2 1 1 1 3 0 2 3 3 1 2 2 1 3 3 0 2 1 2 1 0 1 2 3 1 1 1 2 2 2 0 2 1\n",
      " 0 2 0 2 3 2 1 3 2 1 3 0 1 0 1 1 2 3 0 1 0 2 2 1 3 2 1 0 2 3 3 2 2 1 3 2 0\n",
      " 1 2 1 3 1 2 0 3 3 0 3 0 1 3 0 3 0 1 0 2 3 2 0 1 0 0 1]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABHcAAACPCAYAAABwITd8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x29e7938b080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = np.mean(np.mean(load_params(\"best_X_params.pkl\"),axis = 3), axis = 0)[0:101]\n",
    "X_hot = np.argmax(X, axis = 1)\n",
    "print(X_hot)\n",
    "X_bin = np.zeros((X.shape))\n",
    "X_bin[np.arange(X.shape[0]),X_hot] = 1\n",
    "plt.close('all')\n",
    "plt.figure(figsize = (20,2))\n",
    "plt.subplot(212)\n",
    "plt.imshow(X_bin.T, cmap='hot')\n",
    "ax1 = plt.gca()\n",
    "ax1.set_yticks([])\n",
    "ax1.set_xticklabels([-1,0,20,40,60,80,100], fontdict={'fontsize': 16})\n",
    "plt.subplot(211)\n",
    "plt.imshow(X.T, cmap='hot')\n",
    "ax2 = plt.gca()\n",
    "ax2.set_xticklabels([])\n",
    "ax2.set_yticks([])\n",
    "plt.savefig(\"cs230_generate_X.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  6.58127305e-04  -1.66967620e-05   2.62098183e-04   1.91644242e-04]\n",
      "[ 0.259  0.253  0.249  0.239]\n"
     ]
    }
   ],
   "source": [
    "X = np.mean(np.mean(load_params(\"best_X_params.pkl\"),axis = 3), axis = 0)\n",
    "print(np.mean(X,axis=0))\n",
    "X_hot = np.argmax(X, axis = 1)\n",
    "X_bin = np.zeros((X.shape))\n",
    "X_bin[np.arange(X.shape[0]),X_hot] = 1\n",
    "print(np.mean(X_bin,axis=0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.90033054  0.04042715  0.          0.02745995  0.          0.          0.\n",
      "   0.03178235]\n",
      " [ 0.76190476  0.06051587  0.          0.05257937  0.          0.          0.\n",
      "   0.125     ]\n",
      " [ 0.65637066  0.0965251   0.          0.07722008  0.          0.          0.\n",
      "   0.16988417]\n",
      " [ 0.5751634   0.08496732  0.          0.07843137  0.          0.          0.\n",
      "   0.26143791]\n",
      " [ 0.625       0.10416667  0.          0.04166667  0.          0.          0.\n",
      "   0.22916667]\n",
      " [ 0.39473684  0.07894737  0.          0.02631579  0.          0.          0.\n",
      "   0.5       ]\n",
      " [ 0.47826087  0.11594203  0.          0.07971014  0.          0.          0.\n",
      "   0.32608696]\n",
      " [ 0.1244898   0.02653061  0.          0.04897959  0.          0.          0.\n",
      "   0.8       ]]\n"
     ]
    }
   ],
   "source": [
    "conf_mat_012 = np.array([[7082,  318,    0,  216,    0,    0,    0,  250],\n",
    "       [ 768,   61,    0,   53,    0,    0,    0,  126],\n",
    "       [ 170,   25,    0,   20,    0,    0,    0,   44],\n",
    "       [  88,   13,    0,   12,    0,    0,    0,   40],\n",
    "       [  30,    5,    0,    2,    0,    0,    0,   11],\n",
    "       [  15,    3,    0,    1,    0,    0,    0,   19],\n",
    "       [  66,   16,    0,   11,    0,    0,    0,   45],\n",
    "       [  61,   13,    0,   24,    0,    0,    0,  392]])\n",
    "\n",
    "inputSums = np.repeat(np.sum(conf_mat_012,axis=1,keepdims=True),8,axis = 1)\n",
    "conf_mat_frac_012 = conf_mat_012/inputSums\n",
    "print(conf_mat_frac_012)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}