training.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "source": [
    "!pip install scikit-learn\n",
    "!pip install rasterio\n",
    "!pip install earthpy\n",
    "!pip install gdown\n",
    "from glob import glob\n",
    "import numpy as np\n",
    "from scipy.io import loadmat\n",
    "import rasterio as rio\n",
    "import earthpy as et\n",
    "import earthpy.spatial as es\n",
    "import earthpy.plot as ep\n",
    "import matplotlib.pyplot as plt"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.8/dist-packages (1.0.2)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (3.1.0)\n",
      "Requirement already satisfied: numpy>=1.14.6 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (1.22.1)\n",
      "Requirement already satisfied: scipy>=1.1.0 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (1.8.0)\n",
      "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (1.1.0)\n",
      "\u001B[33mWARNING: You are using pip version 20.2.4; however, version 22.0.3 is available.\n",
      "You should consider upgrading via the '/usr/bin/python3 -m pip install --upgrade pip' command.\u001B[0m\n",
      "Requirement already satisfied: rasterio in /usr/local/lib/python3.8/dist-packages (1.2.10)\n",
      "Requirement already satisfied: click-plugins in /usr/local/lib/python3.8/dist-packages (from rasterio) (1.1.1)\n",
      "Requirement already satisfied: certifi in /usr/lib/python3/dist-packages (from rasterio) (2019.11.28)\n",
      "Requirement already satisfied: snuggs>=1.4.1 in /usr/local/lib/python3.8/dist-packages (from rasterio) (1.4.7)\n",
      "Requirement already satisfied: affine in /usr/local/lib/python3.8/dist-packages (from rasterio) (2.3.0)\n",
      "Requirement already satisfied: cligj>=0.5 in /usr/local/lib/python3.8/dist-packages (from rasterio) (0.7.2)\n",
      "Requirement already satisfied: click>=4.0 in /usr/local/lib/python3.8/dist-packages (from rasterio) (8.0.3)\n",
      "Requirement already satisfied: numpy in /usr/local/lib/python3.8/dist-packages (from rasterio) (1.22.1)\n",
      "Requirement already satisfied: attrs in /usr/local/lib/python3.8/dist-packages (from rasterio) (21.4.0)\n",
      "Requirement already satisfied: setuptools in /usr/local/lib/python3.8/dist-packages (from rasterio) (60.7.0)\n",
      "Requirement already satisfied: pyparsing>=2.1.6 in /usr/local/lib/python3.8/dist-packages (from snuggs>=1.4.1->rasterio) (3.0.7)\n",
      "\u001B[33mWARNING: You are using pip version 20.2.4; however, version 22.0.3 is available.\n",
      "You should consider upgrading via the '/usr/bin/python3 -m pip install --upgrade pip' command.\u001B[0m\n",
      "Requirement already satisfied: earthpy in /usr/local/lib/python3.8/dist-packages (0.9.4)\n",
      "Requirement already satisfied: numpy>=1.14.0 in /usr/local/lib/python3.8/dist-packages (from earthpy) (1.22.1)\n",
      "Requirement already satisfied: rasterio in /usr/local/lib/python3.8/dist-packages (from earthpy) (1.2.10)\n",
      "Requirement already satisfied: matplotlib>=2.0.0 in /usr/local/lib/python3.8/dist-packages (from earthpy) (3.5.1)\n",
      "Requirement already satisfied: geopandas in /usr/local/lib/python3.8/dist-packages (from earthpy) (0.10.2)\n",
      "Requirement already satisfied: requests in /usr/lib/python3/dist-packages (from earthpy) (2.22.0)\n",
      "Requirement already satisfied: scikit-image in /usr/local/lib/python3.8/dist-packages (from earthpy) (0.19.1)\n",
      "Requirement already satisfied: setuptools in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (60.7.0)\n",
      "Requirement already satisfied: click>=4.0 in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (8.0.3)\n",
      "Requirement already satisfied: certifi in /usr/lib/python3/dist-packages (from rasterio->earthpy) (2019.11.28)\n",
      "Requirement already satisfied: affine in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (2.3.0)\n",
      "Requirement already satisfied: attrs in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (21.4.0)\n",
      "Requirement already satisfied: snuggs>=1.4.1 in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (1.4.7)\n",
      "Requirement already satisfied: click-plugins in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (1.1.1)\n",
      "Requirement already satisfied: cligj>=0.5 in /usr/local/lib/python3.8/dist-packages (from rasterio->earthpy) (0.7.2)\n",
      "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (21.3)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (4.29.1)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (9.0.0)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (3.0.7)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (2.8.2)\n",
      "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (0.11.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.0.0->earthpy) (1.3.2)\n",
      "Requirement already satisfied: pyproj>=2.2.0 in /usr/local/lib/python3.8/dist-packages (from geopandas->earthpy) (3.3.0)\n",
      "Requirement already satisfied: pandas>=0.25.0 in /usr/local/lib/python3.8/dist-packages (from geopandas->earthpy) (1.4.0)\n",
      "Requirement already satisfied: fiona>=1.8 in /usr/local/lib/python3.8/dist-packages (from geopandas->earthpy) (1.8.20)\n",
      "Requirement already satisfied: shapely>=1.6 in /usr/local/lib/python3.8/dist-packages (from geopandas->earthpy) (1.8.0)\n",
      "Requirement already satisfied: tifffile>=2019.7.26 in /usr/local/lib/python3.8/dist-packages (from scikit-image->earthpy) (2022.2.2)\n",
      "Requirement already satisfied: PyWavelets>=1.1.1 in /usr/local/lib/python3.8/dist-packages (from scikit-image->earthpy) (1.2.0)\n",
      "Requirement already satisfied: networkx>=2.2 in /usr/local/lib/python3.8/dist-packages (from scikit-image->earthpy) (2.6.3)\n",
      "Requirement already satisfied: scipy>=1.4.1 in /usr/local/lib/python3.8/dist-packages (from scikit-image->earthpy) (1.8.0)\n",
      "Requirement already satisfied: imageio>=2.4.1 in /usr/local/lib/python3.8/dist-packages (from scikit-image->earthpy) (2.14.1)\n",
      "Requirement already satisfied: six>=1.5 in /usr/lib/python3/dist-packages (from python-dateutil>=2.7->matplotlib>=2.0.0->earthpy) (1.14.0)\n",
      "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.25.0->geopandas->earthpy) (2021.3)\n",
      "Requirement already satisfied: munch in /usr/local/lib/python3.8/dist-packages (from fiona>=1.8->geopandas->earthpy) (2.5.0)\n",
      "\u001B[33mWARNING: You are using pip version 20.2.4; however, version 22.0.3 is available.\n",
      "You should consider upgrading via the '/usr/bin/python3 -m pip install --upgrade pip' command.\u001B[0m\n",
      "Requirement already satisfied: gdown in /usr/local/lib/python3.8/dist-packages (4.2.1)\n",
      "Requirement already satisfied: six in /usr/lib/python3/dist-packages (from gdown) (1.14.0)\n",
      "Requirement already satisfied: beautifulsoup4 in /usr/local/lib/python3.8/dist-packages (from gdown) (4.10.0)\n",
      "Requirement already satisfied: filelock in /usr/local/lib/python3.8/dist-packages (from gdown) (3.4.2)\n",
      "Requirement already satisfied: tqdm in /usr/local/lib/python3.8/dist-packages (from gdown) (4.62.3)\n",
      "Requirement already satisfied: requests[socks] in /usr/lib/python3/dist-packages (from gdown) (2.22.0)\n",
      "Requirement already satisfied: soupsieve>1.2 in /usr/local/lib/python3.8/dist-packages (from beautifulsoup4->gdown) (2.3.1)\n",
      "Requirement already satisfied: PySocks!=1.5.7,>=1.5.6 in /usr/local/lib/python3.8/dist-packages (from requests[socks]->gdown) (1.7.1)\n",
      "\u001B[33mWARNING: You are using pip version 20.2.4; however, version 22.0.3 is available.\n",
      "You should consider upgrading via the '/usr/bin/python3 -m pip install --upgrade pip' command.\u001B[0m\n"
     ]
    }
   ],
   "metadata": {
    "collapsed": true,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "source": [
    "!cd ~\n",
    "!gdown --id \"1i5_vAMNI3PqD5nQwUN-zMXaSNo1r1qUm\""
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Downloading...\n",
      "From: https://drive.google.com/uc?id=1i5_vAMNI3PqD5nQwUN-zMXaSNo1r1qUm\n",
      "To: /tf/dataset1.zip\n",
      "100%|████████████████████████████████████████| 122k/122k [00:00<00:00, 2.99MB/s]\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "source": [
    "!ls\n",
    "!unzip ./dataset1.zip"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Pavia_University_Model.h5  dataset1\t datasett      tensorflow-tutorials\n",
      "aspiranturelogs\t\t   dataset1.zip  datasett.zip\n",
      "Archive:  ./dataset1.zip\n",
      "replace dataset1/labels_test.png? [y]es, [n]o, [A]ll, [N]one, [r]ename: ^C\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "source": [
    "files = glob(\"dataset1/*.png\")\n",
    "files.sort()\n",
    "files"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "['dataset1/labels_test.png']"
      ]
     },
     "metadata": {},
     "execution_count": 2
    }
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "source": [
    "import skimage.io as io\n",
    "# l = []\n",
    "# for i in files:\n",
    "#   with rio.open(i, 'r') as f:\n",
    "#     l.append(f.read(0))\n",
    "#     l.append(f.read(1))\n",
    "#     l.append(f.read(2))\n",
    "l = []\n",
    "for i in files:\n",
    "  im = plt.imread(i)\n",
    "  l.append(im[:, :, 0])\n",
    "  l.append(im[:, :, 1])\n",
    "  l.append(im[:, :, 2])\n"
   ],
   "outputs": [],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "source": [
    "#print(l)\n",
    "# plt.imshow(ffd[0])\n",
    "# ff = ffd[0]\n",
    "# fff = ff[: , : , 0]\n",
    "# print(fff.shape)\n",
    "#print(ffd[0,:,:,0])\n",
    "#plt.show()"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "source": [
    "arr_st = np.stack(l)"
   ],
   "outputs": [],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "source": [
    "y_data=loadmat(\"dataset1/matlab.mat\")[\"data\"]"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "source": [
    "plt.matshow(y_data)\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnEAAAGkCAYAAABelnM5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAA9hAAAPYQGoP6dpAAA2e0lEQVR4nO3df3CU9YHH8U8CJBAhCQhJ2AocGIuiCCfWJFfhziNDQI5T5GZEGQ8txSsN3gmWIm1FrNci2NGqg9rOndLOqLTeFD1/oQhCWk2icKYUUM5YavRiggXzA5AQ4Ht/0F2zySbZ3TzPPr/er5mMkn2y+e43m+w73+dH0owxRgAAAPCUdKcHAAAAgMQRcQAAAB5ExAEAAHgQEQcAAOBBRBwAAIAHEXEAAAAeRMQBAAB4EBEHAADgQUQcAACABxFxAAAAHuTLiNuwYYP+6q/+SgMHDlRRUZHefvttp4fkCWvWrFFaWlrU24UXXhi5/cSJEyovL9e5556rwYMHa968eWpsbIy6j7q6Os2ePVtZWVnKy8vTihUrdOrUqVQ/FMdVVFRozpw5CoVCSktL03PPPRd1uzFGq1ev1siRIzVo0CCVlpbqgw8+iNrmyJEjWrBggbKzs5Wbm6tFixbp6NGjUdvs2bNHU6dO1cCBAzVq1CitX7/e7ofmCr3N780339zluTxz5syobZjf2NauXauvfe1rGjJkiPLy8nTttdfqwIEDUdtY9bNgx44duuyyy5SZmanCwkJt3LjR7ofnuHjm9+/+7u+6PH+/9a1vRW3D/Hb12GOP6dJLL1V2drays7NVUlKiV155JXK7L5+3xmc2bdpkMjIyzBNPPGH27dtnFi9ebHJzc01jY6PTQ3O9u+++21x88cXm008/jbx99tlnkdu/9a1vmVGjRplt27aZXbt2meLiYvM3f/M3kdtPnTplLrnkElNaWmreffdd8/LLL5vhw4ebVatWOfFwHPXyyy+b73//++Y3v/mNkWQ2b94cdft9991ncnJyzHPPPWd+//vfm3/8x380Y8eONV988UVkm5kzZ5pJkyaZqqoq89vf/tYUFhaaG264IXJ7c3Ozyc/PNwsWLDB79+41zzzzjBk0aJD52c9+lqqH6Zje5nfhwoVm5syZUc/lI0eORG3D/MZWVlZmnnzySbN3715TU1Njrr76ajN69Ghz9OjRyDZW/Cz44x//aLKysszy5cvN/v37zSOPPGL69etntmzZktLHm2rxzO/f/u3fmsWLF0c9f5ubmyO3M7+x/fd//7d56aWXzP/+7/+aAwcOmO9973tmwIABZu/evcYYfz5vfRdxV1xxhSkvL4/8+/Tp0yYUCpm1a9c6OCpvuPvuu82kSZNi3tbU1GQGDBhgnn322cj73nvvPSPJVFZWGmPOvrCmp6ebhoaGyDaPPfaYyc7ONm1tbbaO3c06R8aZM2dMQUGBuf/++yPva2pqMpmZmeaZZ54xxhizf/9+I8m88847kW1eeeUVk5aWZv7v//7PGGPMo48+aoYOHRo1tytXrjTjx4+3+RG5S3cRd80113T7Mcxv/A4dOmQkmZ07dxpjrPtZ8N3vftdcfPHFUZ/r+uuvN2VlZXY/JFfpPL/GnI24f/u3f+v2Y5jf+A0dOtT8x3/8h2+ft77anXry5Ent3r1bpaWlkfelp6ertLRUlZWVDo7MOz744AOFQiGNGzdOCxYsUF1dnSRp9+7dam9vj5rbCy+8UKNHj47MbWVlpSZOnKj8/PzINmVlZWppadG+fftS+0Bc7ODBg2poaIiay5ycHBUVFUXNZW5uri6//PLINqWlpUpPT1d1dXVkm2nTpikjIyOyTVlZmQ4cOKDPP/88RY/GvXbs2KG8vDyNHz9eS5Ys0eHDhyO3Mb/xa25uliQNGzZMknU/CyorK6PuI7xN0H5Wd57fsKeeekrDhw/XJZdcolWrVun48eOR25jf3p0+fVqbNm3SsWPHVFJS4tvnbX9HPqtN/vznP+v06dNRXwBJys/P1/vvv+/QqLyjqKhIGzdu1Pjx4/Xpp5/qnnvu0dSpU7V37141NDQoIyNDubm5UR+Tn5+vhoYGSVJDQ0PMuQ/fhrPCcxFrrjrOZV5eXtTt/fv317Bhw6K2GTt2bJf7CN82dOhQW8bvBTNnztR1112nsWPH6sMPP9T3vvc9zZo1S5WVlerXrx/zG6czZ87o9ttv19e//nVdcsklkmTZz4LutmlpadEXX3yhQYMG2fGQXCXW/ErSjTfeqDFjxigUCmnPnj1auXKlDhw4oN/85jeSmN+e/OEPf1BJSYlOnDihwYMHa/PmzZowYYJqamp8+bz1VcShb2bNmhX5/0svvVRFRUUaM2aMfv3rX/v2Gx7+NH/+/Mj/T5w4UZdeeqnOP/987dixQ9OnT3dwZN5SXl6uvXv36ne/+53TQ/Gl7ub31ltvjfz/xIkTNXLkSE2fPl0ffvihzj///FQP01PGjx+vmpoaNTc367/+67+0cOFC7dy50+lh2cZXu1OHDx+ufv36dTnbpLGxUQUFBQ6Nyrtyc3P11a9+VbW1tSooKNDJkyfV1NQUtU3HuS0oKIg59+HbcFZ4Lnp6nhYUFOjQoUNRt586dUpHjhxhvpMwbtw4DR8+XLW1tZKY33gsXbpUL774ot544w2dd955kfdb9bOgu22ys7MD8Utjd/MbS1FRkSRFPX+Z39gyMjJUWFioKVOmaO3atZo0aZIeeugh3z5vfRVxGRkZmjJlirZt2xZ535kzZ7Rt2zaVlJQ4ODJvOnr0qD788EONHDlSU6ZM0YABA6Lm9sCBA6qrq4vMbUlJif7whz9EvThu3bpV2dnZmjBhQsrH71Zjx45VQUFB1Fy2tLSouro6ai6bmpq0e/fuyDbbt2/XmTNnIj/QS0pKVFFRofb29sg2W7du1fjx4wOxqy8Rn3zyiQ4fPqyRI0dKYn57YozR0qVLtXnzZm3fvr3LLmWrfhaUlJRE3Ud4G7//rO5tfmOpqamRpKjnL/MbnzNnzqitrc2/z1tHTqew0aZNm0xmZqbZuHGj2b9/v7n11ltNbm5u1NkmiO2OO+4wO3bsMAcPHjRvvvmmKS0tNcOHDzeHDh0yxpw9PXv06NFm+/btZteuXaakpMSUlJREPj58evaMGTNMTU2N2bJlixkxYkQgLzHS2tpq3n33XfPuu+8aSeaBBx4w7777rvnoo4+MMWcvMZKbm2uef/55s2fPHnPNNdfEvMTIX//1X5vq6mrzu9/9zlxwwQVRl8Boamoy+fn55qabbjJ79+41mzZtMllZWb6/BIYxPc9va2ur+c53vmMqKyvNwYMHzeuvv24uu+wyc8EFF5gTJ05E7oP5jW3JkiUmJyfH7NixI+oSF8ePH49sY8XPgvClGlasWGHee+89s2HDBt9fAsOY3ue3trbW/PCHPzS7du0yBw8eNM8//7wZN26cmTZtWuQ+mN/Y7rzzTrNz505z8OBBs2fPHnPnnXeatLQ089prrxlj/Pm89V3EGWPMI488YkaPHm0yMjLMFVdcYaqqqpwekidcf/31ZuTIkSYjI8N85StfMddff72pra2N3P7FF1+Yb3/722bo0KEmKyvLzJ0713z66adR9/GnP/3JzJo1ywwaNMgMHz7c3HHHHaa9vT3VD8Vxb7zxhpHU5W3hwoXGmLOXGbnrrrtMfn6+yczMNNOnTzcHDhyIuo/Dhw+bG264wQwePNhkZ2ebW265xbS2tkZt8/vf/95ceeWVJjMz03zlK18x9913X6oeoqN6mt/jx4+bGTNmmBEjRpgBAwaYMWPGmMWLF3f5RY75jS3WvEoyTz75ZGQbq34WvPHGG2by5MkmIyPDjBs3Lupz+FVv81tXV2emTZtmhg0bZjIzM01hYaFZsWJF1HXijGF+Y/nGN75hxowZYzIyMsyIESPM9OnTIwFnjD+ft2nGGJO6dT8AAABYwVfHxAEAAAQFEQcAAOBBRBwAAIAHEXEAAAAeRMQBAAB4EBEHAADgQb6NuLa2Nq1Zs0ZtbW1OD8WXmF/7MLf2Yn7txfzah7m1lxfn19XXiduwYYPuv/9+NTQ0aNKkSXrkkUd0xRVXxPWxLS0tysnJUXNzs7Kzs20eafAwv/Zhbu3F/NqL+bUPc2svL86va1fifvWrX2n58uW6++679T//8z+aNGmSysrKuvzRagAAgCBybcQ98MADWrx4sW655RZNmDBBjz/+uLKysvTEE084PTQAAADH9Xd6ALGcPHlSu3fv1qpVqyLvS09PV2lpqSorK2N+TFtbW9R+7KamJklSc3OzrWMNqpaWlqj/wjrMrb2YX3sxv/Zhbu3llvk1xqi1tVWhUEjp6T2vtbky4v785z/r9OnTys/Pj3p/fn6+3n///Zgfs3btWt1zzz1d3j969GhbxoizRo0a5fQQfIu5tRfzay/m1z7Mrb3cMr8ff/yxzjvvvB63cWXEJWPVqlVavnx55N/Nzc0aPXq0psz8vvoNGOjgyJCohq+nOT0EIKZxd77j9BDiVrBtcMo/5x9/emHKPyfgN6fbT2j3lh9pyJAhvW7ryogbPny4+vXrp8bGxqj3NzY2qqCgIObHZGZmKjMzs8v7+w0YqP5EnKekDyTi4E790wZYdl+hqt5/QHtJfXGr+s/lZy1glbS03l8LXRlxGRkZmjJlirZt26Zrr71WknTmzBlt27ZNS5cudXZwAALLb+EFwNtcGXGStHz5ci1cuFCXX365rrjiCv30pz/VsWPHdMsttzg9NAABM61kn9NDAIAuXBtx119/vT777DOtXr1aDQ0Nmjx5srZs2dLlZAcAAIAgcu114iRp6dKl+uijj9TW1qbq6moVFRU5PSQAAVRRebHTQwCALlwdcQAAAIiNiAMAAPAgIg4AAMCDXHtiAwAAcIeszdVODyEwTpn2uLcl4gAAsAChg1Qj4gAAfVJf3Or0EFyj9sFip4eQlMJlVU4PAUkg4gAACDivxqcfnTlxQrrz+bi25cQGAAAADyLiAAAAPIiIA4A48FcbALgNEQcAAOBBRBwAxImzMAG4CREHAHEIX4KBkAPgFkQcACSIkAPgBkQcACSBkAPgNCIOAJJEyAFwEhEHAH0Q9JAL+uMHnETEAUAfETIAnEDEAYAFCDkAqUbEAYBFCDkAqUTEAYCFCDkAqULEAYDFCDkAqUDEAYANCDkAdiPiAMAmhBwAOxFxAGAjQg6AXYg4ALAZIQfADkQcAKQAIQfAakQcAKQIIQfASkQcAKQQIQfAKkQcAKSYX0LOL48D8CoiDgAAwIOIOABwQH1xq+9WsrI2Vzs9BCBQiDgAcJDfQg5A6hBxAAAAHkTEAYDDWI0DkAwiDgBcgJADkCgiDgBcgpADkAgiDgBchJADEC8iDgBchpADEA8iDgBciJAD0BsiDgBcipAD0BMiDgBcjJAD0B0iDgBcjpADEAsRBwAe4LaQc9t4gCAi4gDAIwgnAB0RcQDgIYQcgDAiDgA8hpADIBFxAOBJhBwAIg4APIqQc5fjc4ucHgIChogDAA8j5IDgIuIAwOMIOSCYiDgA8AFCDggeIg4AfIKQA4KFiAMAHyHkgOAg4gAAADyIiAMAn6kvbmVFDggAIg4AfMqukCMQAXcg4gAAADyIiAOAXhQuq3J6CElj1Qzwr/5OD8BuWS/sUv+0AU4PA4mYVuz0CABfqS9uVahqiNPDAGAx30ecJM/88OI3ZgB2IeQA/wlExHlFkH/AErCA/Qg5wF84Jg6uwAsLkBr8wgT4BxEHAAFDyAH+QMQBQAARcoD3EXEAEFCEHOBtRBwABBghB3gXEQcAAZdIyBF9gHsQcQAA4gzwICIOACCJkAO8hogDAEQQcoB3EHEAgCiEHOANRBwAoAtCDnA/Ig4AEBMhB7gbEQcA6BYhB7gXEQcA6BEhB7gTEQcA6BUhB7iP5RG3Zs0apaWlRb1deOGFkdtPnDih8vJynXvuuRo8eLDmzZunxsbGqPuoq6vT7NmzlZWVpby8PK1YsUKnTp2yeqgAgAQQcoC79LfjTi+++GK9/vrrX36S/l9+mmXLlumll17Ss88+q5ycHC1dulTXXXed3nzzTUnS6dOnNXv2bBUUFOitt97Sp59+qn/+53/WgAED9OMf/9iO4QIAAHiOLRHXv39/FRQUdHl/c3Oz/vM//1NPP/20/v7v/16S9OSTT+qiiy5SVVWViouL9dprr2n//v16/fXXlZ+fr8mTJ+vee+/VypUrtWbNGmVkZNgxZAAAAE+x5Zi4Dz74QKFQSOPGjdOCBQtUV1cnSdq9e7fa29tVWloa2fbCCy/U6NGjVVlZKUmqrKzUxIkTlZ+fH9mmrKxMLS0t2rdvX7efs62tTS0tLVFvAAAAfmV5xBUVFWnjxo3asmWLHnvsMR08eFBTp05Va2urGhoalJGRodzc3KiPyc/PV0NDgySpoaEhKuDCt4dv687atWuVk5MTeRs1apS1DwwAAMBFLN+dOmvWrMj/X3rppSoqKtKYMWP061//WoMGDbL600WsWrVKy5cvj/y7paWFkAMAAL5l+yVGcnNz9dWvflW1tbUqKCjQyZMn1dTUFLVNY2Nj5Bi6goKCLmerhv8d6zi7sMzMTGVnZ0e9AQAA+JUtJzZ0dPToUX344Ye66aabNGXKFA0YMEDbtm3TvHnzJEkHDhxQXV2dSkpKJEklJSX60Y9+pEOHDikvL0+StHXrVmVnZ2vChAl2DxcAkGJZm6ttu+/jc4tsu2/AaZZH3He+8x3NmTNHY8aMUX19ve6++27169dPN9xwg3JycrRo0SItX75cw4YNU3Z2tm677TaVlJSouLhYkjRjxgxNmDBBN910k9avX6+Ghgb94Ac/UHl5uTIzM60eLgDAQXYGXCruv4tpxan9fAg0yyPuk08+0Q033KDDhw9rxIgRuvLKK1VVVaURI0ZIkh588EGlp6dr3rx5amtrU1lZmR599NHIx/fr108vvviilixZopKSEp1zzjlauHChfvjDH1o9VACAg1IeWIDPWB5xmzZt6vH2gQMHasOGDdqwYUO324wZM0Yvv/yy1UMDALhErIDrvOuTyAN6ZvsxcQAAdBRPwIXf13nb2gft3V1ZuKzK1vsHrETEAQBSpnOUdYy3+mlpkqRQhelye/jjCpdV2RpydkciYCXbLzECAEBP6qelRQKuOx1jr3BZFStmgIg4AECKxNqNGiveugu643OLusQcEGREHADAdrECLtldl6zKAWcRcQCAlOst4OLZvcqqHIKOiAMA2CZrc7WtZ5iyKocgI+IAALbI2lzd5dIhiQRcb6txYazKIaiIOACA5foacMnoGHOsyiEIiDgAgKU6Blxf/+pCvKtxHbEqh6Ag4gAAvsOqHIKAiAMAWKa7Vbi+7EpNZjUujFU5+BkRBwCwTKy/gWrFsXB9DTlW5eBHRBwAuFA4PGJFkVt1HGtfj4WzA6ty8Jv+Tg8AANCzZEPODSFl5Rmp9dPSFKowfbqPjrt6wyHHH72HVxFxAOBTqYw/t6/CdXZ8blFknIXLqgg5eBIRBwA9cGq3W6zrrKWKVZ/XjjCyYjUujFU5eB3HxAEAAo1j5eBVRBwAwDLhXZR2rmj15UzV7nAGK7yIiAMAeI4dISd1jTnAzYg4AIAlUrEKlyqEHLwgECc21Be3Oj2EhL1aX2PJ/dxSN9WS+wEAt7HyJIdYOoecH+IU/hKIiAuyJ0f/1pHPWxaa3OV9oaoh3W7vxdAG7ObkGaqJ8tMqXGfhy5EQc3AbIq4bPQVHPIIeJZ1XEstCkwM/JwCsZ/dqXFjny5EQcnADjolDSli1exiA+/h5Fa4zjpWDm7ASBwDwtFStxoVxrBzcgpU4F2LVCoBXBGkVrjNW5eA0Iq4bHL8FwA288HdI3cCu68b1JnxdOS4QDCcQcUgZVhgBfwriKlxnrMrBCRwTBwBIittWCVN9bFxnHCuHVGMl7i/6ekkRq7Baxd8tBLyEUOmKVTmkCitx+jLgQlVDOBbORfgBmBxeVJEKbluFC3N6NS6MVTmkAhHXSceQqy9uTXqFjiD0vu6+9m7/uhK/yXHzi6xb/3KDW+fMLSEnRcecW+cL3kXExUCABVtv4d7xdr8+T9wYDHYjfuPn1lU4NwufwSq5N37hPURcNwg5d3HLMYudBSHogiKI4doXbg8RN63GhbEqB6sF/sSGnuKAkLOen0/cCFUNibx5HSst6A7Pjb7junKwCitxvUj1C3JZaLKvQycovBxy4V9c3HocFpznlVUkN67GhbEqBysEfiUOABAfL67CuX3MHY+VAxJFxAEA4ua1VaPaB4s9EXKhCkPMIWFEHIBuuf3FD6nDc8F+HCuHRBFxAIC4eG0VLswLq3Fhx+cWsYsVcQt0xLn14POy0GSnh+A6nCXsHK+8+ME+fnkOeOlxEHKIR6AjDkB8vPTiB3t4dRUuzIvj51g59IaIAwB0y08B76Xdqh0Rc+gOEYeUYlexd3nxxQ/W8OIqVk+8+lxmFys6I+IAAIHh9SANr8oBUoAjzq0nNQBu5tUVDCQna3O156OnO15/LrN7FVKAI87t2O3YFWeoAqnj9cjpSThMvf4YWZUDEYeUIUz9wesvfIifX1fhOvLD85lVueAi4lyM6AHgBD+ETW/8FqisygVTICOO4+GAvgnCi3zQ+S1yeuKn5zOrcsESyIhD6lmxqkh8A7BSx1D1U8ixKhcc/Z0egJ+Fqob0+WD8stBkvVpfY82AXI5I85aszdU6PrfI6WHAYn4+IzVowiFXPy3N4ZHALqzEwXGhqiEEnEf5afUCweTX1biOWJXzLyLOA7x+goPXxw8EBatw/g45jpXzHyIOQJ/49UUvaIL8dewcrn6dC46V85/ARRy77QDr+fVFL2iCvAoXpMceqjDEnE8ELuK8yqu7JL06bgDBFoRfTAg57yPiPMRLQVQWmuyp8aLvgvCi51ccC3dWEOeAVTlvI+I8xgth5IUxwh6EHPwkSM9nQs6biDgPcnMkuXlsALpiFS5aUE5yiIVVOe8JVMRxUoN92H2KsCC96HkdX6v4BG2eCDnvCFTE+YmbgslNY4E7BO1Fz8tYheuKOWFVziuIOA9zOp5YfUNPCDl34+uTmKDOFyHnbkSczezehetURBFvgPex4tS9WHNDyMFtiLgU8FPIsfqGRAT1RQ/+dHxukdNDcAy7V90pMBHn95MaUhFXdty/378ugBtxRmp8Os5RkAOuI0LOXfo7PYCgCFUNUX1xq+2fJxxar9bXWH6fQDKyNlfzAugirI7Gp+Mfi+f5Gy0ccvXT0hweCYi4FEpVyEmJxxyhBjsRcu7CKlxXHaOtI5633QtVGELOYURciqUy5CTiDO5ByMFNuos2iXBLBKtyziLiACAggnwsHNFmL1blnBGIiHPbwfOpXo0D3ILVOKQS4ZZarMqlXiAizo0IOQQVIeeMIKzCEW3uwKpc6hBxDiLkAKBvCDd3YlUuNYg4hxFyCCJW41LLT6twPUWbRLi5Daty9iLiADiCkEO8WG3zNlbl7OP7iCvYNtjpIfSK1TgAdvHiKhzR5k+sylkv4T+7VVFRoTlz5igUCiktLU3PPfdc1O3GGK1evVojR47UoEGDVFpaqg8++CBqmyNHjmjBggXKzs5Wbm6uFi1apKNHj0Zts2fPHk2dOlUDBw7UqFGjtH79+sQfnYe47QxaIBX46wEIK1xWFXnr7PjcosgbvI0/22WthCPu2LFjmjRpkjZs2BDz9vXr1+vhhx/W448/rurqap1zzjkqKyvTiRMnItssWLBA+/bt09atW/Xiiy+qoqJCt956a+T2lpYWzZgxQ2PGjNHu3bt1//33a82aNfr5z3+exEP0DkIOQUTI2cfNq3Ado41wC5ZQhSHmLJLw7tRZs2Zp1qxZMW8zxuinP/2pfvCDH+iaa66RJP3yl79Ufn6+nnvuOc2fP1/vvfeetmzZonfeeUeXX365JOmRRx7R1VdfrZ/85CcKhUJ66qmndPLkST3xxBPKyMjQxRdfrJqaGj3wwANRsQcA8AZOSEBnHCvXd5YeE3fw4EE1NDSotLQ08r6cnBwVFRWpsrJS8+fPV2VlpXJzcyMBJ0mlpaVKT09XdXW15s6dq8rKSk2bNk0ZGRmRbcrKyrRu3Tp9/vnnGjp0aJfP3dbWpra2tsi/W1parHxoKcPxcQgiTnKwntOrcEQb4sWxcsmzNOIaGhokSfn5+VHvz8/Pj9zW0NCgvLy86EH0769hw4ZFbTN27Ngu9xG+LVbErV27Vvfcc481D8RhhByCiJDzPsINySLkkuObs1NXrVql5cuXR/7d0tKiUaNGOTiiviHkACQrVatwRBusxO7VxFkacQUFBZKkxsZGjRw5MvL+xsZGTZ48ObLNoUOHoj7u1KlTOnLkSOTjCwoK1NjYGLVN+N/hbTrLzMxUZmamJY/DLQg5BA2rce5HuMFuxFz8LI24sWPHqqCgQNu2bYtEW0tLi6qrq7VkyRJJUklJiZqamrR7925NmTJFkrR9+3adOXNGRUVFkW2+//3vq729XQMGDJAkbd26VePHj4+5KxWAfxByfWP1Klxv0SYRbrAHu1h7l3DEHT16VLW1tZF/Hzx4UDU1NRo2bJhGjx6t22+/Xf/+7/+uCy64QGPHjtVdd92lUCika6+9VpJ00UUXaebMmVq8eLEef/xxtbe3a+nSpZo/f75CoZAk6cYbb9Q999yjRYsWaeXKldq7d68eeughPfjgg9Y8ag9hNQ5AqrHa5h3xXqLHq18zQq5nCUfcrl27dNVVV0X+HT4ObeHChdq4caO++93v6tixY7r11lvV1NSkK6+8Ulu2bNHAgQMjH/PUU09p6dKlmj59utLT0zVv3jw9/PDDkdtzcnL02muvqby8XFOmTNHw4cO1evXqwF5ehJBD0LAal5xkV+GINmdxrcSesXu1e2nGGF9eca+lpUU5OTmav22BMgZn9P4BHuDHkOMCx+7jpucZ8RC/cAjEE3HsIrWe20PMT19Pv8fcmRMnVHfnD9Tc3Kzs7Owet/XN2akA/IcVucT0FHCEW2LcHmVBxi7WL/k+4hqmH9Xo6mFOD8MS7FYF0J3OAUe0JYZo8xZC7izfR5zfEHIIGlbjepe1uVqFm3vfjnmEn3CsnJTu9ACQOI4jQ9CwSpI8/pA8/C4cc0EUiJW4+uJW34UPK3IApK6BS6wlLgi/JPh9RdtPq3Lj7nxHdXFuG4iI8ytCDkHi9xehZITjg3lJPa8sDATtNcIPx8odn3O59MLzcW1LxAHwDEIuGnORvMKV+7/8x0pvBBni4+VVucJlVWr5h0lxbx+YY+L8+tuIV34bBAAglbx4rFyiv5gFJuL8jJBDkATh+CUA1ghVGM/EXDyXBeqMiPMJL4acF8cMdyDkACTCCyGXzOERRBwAAD7FL8tfcvOqXDKrcFLAIs6vx8WF8c2KIGE1DkAy3BhyyZ6kFKiICwJCDkFCyCFZtesmOD0EOMjNq3KJIOJ8iJADAKB3bgi5ZHelSgGMOL/vUg0j5BAUrMYB8eF7JTanQ64v13sMXMQB8B9enAD0hVO7V/uyCicRcb7GahyChJAD0Fepjrm+/tWVQEZcUHapSoQcAACJSkXIWfE5AhlxQUPIIShYjQNgFaePlYsHEQfAVwg5AFaxa/eqVfcZ2IgL0i5VidU4AACS5dZVucBGXBARcggKVuMAWM2NFwgm4gKGkENQEHJAtL6eCYmz+hpyVoYgERdAhBwAAMlzy6pcoCMuaMfFAUHDahwAO4VjLt6oszr8Ah1xQcZqHIKCkAOQKokEnRWIuAAj5AAAsEeowqhwWVXkT2vZEXaBj7ig71Il5BAErMYBcMLxuUU6Preoz38jtTuBjzgAwUDIAXCKXWcGE3FgNQ6BQcgB8JP+Tg/ADeqLWwMfMqGqISnftWzl5wv61w8AuuPEz3ekBhGHCC9/ozs9biLSO7I2V3PRUwC+4PuIOz7ncvUfMLDX7WrXJXf/hSv3J/eBLhWqGqLadROcHobnJPv8sZJVz0W/B07W5mpCDoAv+D7i7EbwwC14Lsbn+NwiQg6AL3BiA4DACccbJzoA8DIiDkAgEXIAvI6IAxBYhBwALyPiAAQaIQfAq4g4AIHHCQ4AvIiIAwB9edYqAHgFEQcAf0HIAfASIg4AOiDkAHgFEQcAnRByALyAiAOAGAg5AG5HxAFANwg5AG5GxAFADwg5AG5FxAFALwg5eF2oaojTQ4ANiDgAiAMXBPYXohx+QMQBAAB4EBEHAADgQUQcAACABxFxAAAAHkTEAQAAeBARBwAA4EFEHAAAgAcRcQAAAB5ExAEAAHgQEQcAAOBBRBwAAIAHEXEAAAAeRMQBAAKpvrjV6SEAfULEAQAAeBARBwAIlKzN1U4PAbAEEQcAAOBBRBwAALAFq5726u/0AAAAgP0KV+5X7boJKf+8WZurdXxuUco/r13iDdNUPGYiDgAQWEE6Q9VPIdWdoK38EXEAgMAI2ot8mBsDLqhfCysRcQAAwHIdI41gswcRBwAIhKCEhBtX3fyk9sHiuLYLVRibR0LEAQDgKKLLOvEGll8QcQAA37NrFY4A65ugRZfViDgAgO8RW84j2KzHxX4BAIDlghzOhcuqUvJ5iDgAAAAPIuIAAAA8KOGIq6io0Jw5cxQKhZSWlqbnnnsu6vabb75ZaWlpUW8zZ86M2ubIkSNasGCBsrOzlZubq0WLFuno0aNR2+zZs0dTp07VwIEDNWrUKK1fvz7xRwcAAOBTCUfcsWPHNGnSJG3YsKHbbWbOnKlPP/008vbMM89E3b5gwQLt27dPW7du1YsvvqiKigrdeuutkdtbWlo0Y8YMjRkzRrt379b999+vNWvW6Oc//3miwwUAAPClhM9OnTVrlmbNmtXjNpmZmSooKIh523vvvactW7bonXfe0eWXXy5JeuSRR3T11VfrJz/5iUKhkJ566imdPHlSTzzxhDIyMnTxxRerpqZGDzzwQFTsAQAABJUtx8Tt2LFDeXl5Gj9+vJYsWaLDhw9HbqusrFRubm4k4CSptLRU6enpqq6ujmwzbdo0ZWRkRLYpKyvTgQMH9Pnnn8f8nG1tbWppaYl6AwAA8CvLI27mzJn65S9/qW3btmndunXauXOnZs2apdOnT0uSGhoalJeXF/Ux/fv317Bhw9TQ0BDZJj8/P2qb8L/D23S2du1a5eTkRN5GjRpl9UMDAABwDcsv9jt//vzI/0+cOFGXXnqpzj//fO3YsUPTp0+3+tNFrFq1SsuXL4/8u6WlhZADAAC+ZfslRsaNG6fhw4ertrZWklRQUKBDhw5FbXPq1CkdOXIkchxdQUGBGhsbo7YJ/7u7Y+0yMzOVnZ0d9QYAAOBXtkfcJ598osOHD2vkyJGSpJKSEjU1NWn37t2RbbZv364zZ86oqKgosk1FRYXa29sj22zdulXjx4/X0KFD7R4yAACA6yUccUePHlVNTY1qamokSQcPHlRNTY3q6up09OhRrVixQlVVVfrTn/6kbdu26ZprrlFhYaHKysokSRdddJFmzpypxYsX6+2339abb76ppUuXav78+QqFQpKkG2+8URkZGVq0aJH27dunX/3qV3rooYeidpcCAAAEWcLHxO3atUtXXXVV5N/hsFq4cKEee+wx7dmzR7/4xS/U1NSkUCikGTNm6N5771VmZmbkY5566iktXbpU06dPV3p6uubNm6eHH344cntOTo5ee+01lZeXa8qUKRo+fLhWr17N5UVSIGtzdbe3haqGxH0/tesmWDEcAADQjTRjjHF6EHZoaWlRTk6Orphzr/oPGOj0cDzHqpjzOmIUAPom/HpS+2CxwyNJncJlVTo+tyipjz3VfkJvv3CXmpubez2+3/KzU+EP4SdfrJirL26VFIyYK1y53+khWIIYBQD/IeLQo95iLggh5wdeidHwLwiSkv4tFgCCgohDXLqLuSCtysF+0c8jb4RndzoGaRhhCsBKRBwSQswB8Yn9vWBNmHYOROIQCCbbrxMHf+ruRSPW6gOAvqsvbo28ddbTiUgA/IuVOCSNVTnAXon8UpS1uZoVOSBgWIlDnx2fWxTzxYNVOSBxPa24ddb5+44VOSBYWImDZWKtzLEqB8Qn3l96OoZb52hjJQ4IFiIOliPmgPgkE27J3A7An4g42IaYA7pKZjdpLOHvKwIOCC4iDrbrLuYIOQSFVeEWRsABkIg4pNDxuUWsyiEwrNpV2hkB514df77x9UEqEHFIKVbl4GdWr7jZeR+wRndnBBPbSAUiDo7oHHOsysGrUhVuEteCc4tELuVCzMFORBwcRczBq1IZbxIBZ6eeouzV+pqu79zQ/X2VhSZ3+zn4+sFqRBxcgZiDF/QWbna9SHMRX/t0N7cx4y0OnT+uY9SxKgerEXFwlVgxR8jBSU6Fm1Ofx6uSWemKFXDJxlt3Ot5fOOiIOViFiIMrdTyTlVU5pJpbwk1iN5xdUhFwPd1/WWgyZ7Oiz4g4uBarckglN4VbGAFnPSfiLRaCDlYg4uB6HWOOVTlYyY3hFkbAJa63OesccE7EWyzhcXTe3SoRdOgZEQfPIOZgBTeHG5LX+WLinXW8zS3x1tmr9TVdzm4l5tETIg6e0znmCDnEo6d4c9uLJC/c1vJCwIURckgEEQfPCv9Qqy8++wOamENnXgq3MF6wreOleOuo8+5ViecFYiPi4HnEHDrqLty88ALIC3XfhefQqwHXUaxj5Xh+oKN0pwcAWOXLmIvvD4/DP+qLWyNvnR2fW+SZFz6vjNPt/HZxZK9GKOzHShx8hVW54PDirtKesMpiD78EUKxj5QAiDr5EzPmTl3eV9oSAs4dfAi7sbMh56/nut1VRtyHi4GvEnPf5NdzCCDh7+C3ggFiIOARCx5gj5Lyhu+Pb/MaPj8kJfjiRIR6v1tdoajnPGZxFxCFQjs8tUu26sz/wiTn3CUq4wVpBCbiw3274maaW/4vTw4ALEHEIJGLOPQg3JMutf0YLSBUiDoF2fG4Ru1gdQLihr4IecKzGQSLiAFblUoRwg12CFnBAGBEH/AUxZz3CDXYLcsCxGgciDugkHHOFK/c7PRTP6hxvhBtgDy5R4z6Fy6pS9rmIOKAbtesmSCLm4kW4IdWCvAoXxiVHgo2IA3pRu24Cu1i7QbjBKQTcl9itGlxEHBAHdrF+qWO4EW2AO7BbNZiIOCABQd3Fyoob3OS3G37m9BBch92qwUTEAUkIQswRboC3xFqNY4XO34g4oA9q103wVcixqxTwrlfra1QWiv7e5fvY34g4oI/8sCpHvMEr2JUKfImIAyzitZgLh9vxuUXSXIcHA8Ay7EINDiIOsJibd7ESbgBS6fjcoi5/5xbWIeIAG7hpVY5wg1+wK7V3Z4+Lm+z0MJAiRBxgI6dijnCDH5WFJnORX3hGKnZrE3FACqRiFyvhBgDBQsQBKWLHqhzhhqDguCqgKyIOSDGrYq523QTCDQACLN3pAQBBFY65ZLnhpAkAgHOIOMBBtesm9CnmClfuJ+bge+xKTQwnfwQHu1MBB1kVYOH76evqHgDAO4g4IMXsXDkrXLmfkAOAgCDigBRI5S5PVuUAuFHhsirVPljs9DB8hYgDbOL0sWrEHAA34E9v2YcTGwALhU80cDrgOnLTWACkBn+iLBhYiQP6wCuBxKocAPgPEQckyCvhFgsnPgDBMLX8X5weAlKAiAPi4OVw64xVOXhZWWgy10ED/oKIA7rhp3CLhZiDV3BgfGLKQpP5k3wBQcQBf5G1uVqhqiFODyPl2MUK+MvxuUVODwEpQsQh0KLCbWXwAi6MVTm4Hatx8WEVLliIOARK+EWAcIuNVTkA8A4iDr5GtCWOVTm4FatxPSsLTXZ6CEgxIg6+0vEHfKhqCNHWB8Qc3Ojs8V41Tg8DcAUiDp7XJdxgKXaxAu4XXoXjpIZgIeLgOZ13pxBu9mNVDm7CteK8JxyXoQrj8Ej8hYiD68U6BoZwcwYxBzc4PrdIZSERcgi8dKcHAMSStbk68hYWqhoSeYOz/H4hZHgDB/Kfxa5Ud0nl14GVOLhCd2ecEWzuxaocnMSZqgARBwexm9QfClfuV31xK6sAcETQj49jFc6dUvX1IOKQMqy2+dfZryExh9QKr8YFPeQSEf45zPepPxBxsBXhFizEHJwSxJDryypc1uZqvkd9gIiDpXo6RoVwC45Q1RDVF/MbP+zX8di4IIZcsvi+9AciDn3S24HFhFtwhb/2xBzsFsSQKwtNTup7iu9DfyHikDDCDYk4uyrXyrE4sFXQQo7vI0hEHOJAtKGvvlyVI+Zgn6CE3NTyf3F6CHCJhC72u3btWn3ta1/TkCFDlJeXp2uvvVYHDhyI2ubEiRMqLy/Xueeeq8GDB2vevHlqbGyM2qaurk6zZ89WVlaW8vLytGLFCp06dSpqmx07duiyyy5TZmamCgsLtXHjxuQeIRLW8UK7PZ2YwIV3kaiOz5menl9Asjr+csDFgOF3CUXczp07VV5erqqqKm3dulXt7e2aMWOGjh07Ftlm2bJleuGFF/Tss89q586dqq+v13XXXRe5/fTp05o9e7ZOnjypt956S7/4xS+0ceNGrV69OrLNwYMHNXv2bF111VWqqanR7bffrm9+85t69dVXLXjIiKW3aJMIN1in43OImIPV/BxyrMKhozRjTNJ/jfazzz5TXl6edu7cqWnTpqm5uVkjRozQ008/rX/6p3+SJL3//vu66KKLVFlZqeLiYr3yyiv6h3/4B9XX1ys/P1+S9Pjjj2vlypX67LPPlJGRoZUrV+qll17S3r17I59r/vz5ampq0pYtW+IaW0tLi3JycnTFnHvVf8DAZB+ib8X7okmwwW71xa1R/2Y3K6zS8eecX3atEnH+d6r9hN5+4S41NzcrOzu7x2379LdTm5ubJUnDhg2TJO3evVvt7e0qLS2NbHPhhRdq9OjRqqyslCRVVlZq4sSJkYCTpLKyMrW0tGjfvn2RbTreR3ib8H3E0tbWppaWlqg3fCmeXaRhrLghlTo/11iZg1U6r8h5fVWOgENnSZ/YcObMGd1+++36+te/rksuuUSS1NDQoIyMDOXm5kZtm5+fr4aGhsg2HQMufHv4tp62aWlp0RdffKFBgwZ1Gc/atWt1zz33JPtwfCeRF0FiDW4QPos1rONzmNU5JKvz31j18wkPCJ6kV+LKy8u1d+9ebdq0ycrxJG3VqlVqbm6OvH388cdODynl4l1pC2O1DW7T3XOS1Tn0RedfAry4IscqHGJJaiVu6dKlevHFF1VRUaHzzjsv8v6CggKdPHlSTU1NUatxjY2NKigoiGzz9ttvR91f+OzVjtt0PqO1sbFR2dnZMVfhJCkzM1OZmZnJPBzPSvRFjWCDV3RelQvj8iRIlpdX5Ag4dCehiDPG6LbbbtPmzZu1Y8cOjR07Nur2KVOmaMCAAdq2bZvmzZsnSTpw4IDq6upUUlIiSSopKdGPfvQjHTp0SHl5eZKkrVu3Kjs7WxMmTIhs8/LLL0fd99atWyP3EVTJrEQQbvCqjteW64yYQzJihZzk3pMeiDf0JqGzU7/97W/r6aef1vPPP6/x48dH3p+TkxNZIVuyZIlefvllbdy4UdnZ2brtttskSW+99Zaks5cYmTx5skKhkNavX6+GhgbddNNN+uY3v6kf//jHks5eYuSSSy5ReXm5vvGNb2j79u3613/9V7300ksqKyuLa6x+OTuVcAPOihVzYcQcEhHr56qbQo54C7ZEzk5NKOLS0tJivv/JJ5/UzTffLOnsxX7vuOMOPfPMM2pra1NZWZkeffTRyK5SSfroo4+0ZMkS7dixQ+ecc44WLlyo++67T/37f7kwuGPHDi1btkz79+/Xeeedp7vuuivyOeLh1YhL9rgfwg1B0FPIScQc4tPdz1k3hBwBB9sizku8EnF9OVibcENQsSoHu2Rtrk55zBFu6CiRiONvpzqAcAP6prsTH6Sz31+EHPqiu7NXOz+vfrvhZ3HdH5EGuxBxKdDXSyMQbvCr3naPAqnW+eSHnn4hIM7gNCLOJqy2wY/cHF2svsEqnUMOcCsiziKstsGt3BJeRBa8hJCDFxBxfUC4IRWcijCiC+AYS7gbEZcAK34rI9yCIRXhxQsLYC9W4+B2RFwPrPrmJdzcz+roIrAAfwiHHN/TcCMirhPCzZv6GmH8gAbQHX4+wK0CH3FWLpUTbvFLNLp6/SE6tw+DAQDAgwIXcURb8iy9Sj7RBQBAn/g+4rJe2KX+aQMsuz+vhVs4vCzZHUB4AQmx8pdGdukB6Mz3EWcFp8Ktdt2Evt8J4eUIzmiDVYg3AN0h4rqRbLhZeZZjlggBIMgIOFgh3j8jBuuE59zu+fZtxBljJEmn1C6ZxD++ruiIxSMCgMRk/OZ3Tg/B9Y7PudzpIcQl64Vdjn3uUx3/v/2EY+MIklOmXVLP38PdPXdP/+VrFO6YnqSZeLbyoD/+8Y86//zznR4GAABAwj7++GOdd955PW7j25W4YcOGSZLq6uqUk5Pj8Gj8p6WlRaNGjdLHH3+s7Oxsp4fjK8ytvZhfezG/9mFu7eWW+TXGqLW1VaFQqNdtfRtx6enpkqScnBye7DbKzs5mfm3C3NqL+bUX82sf5tZebpjfeBef0m0eBwAAAGxAxAEAAHiQbyMuMzNTd999tzIzM50eii8xv/Zhbu3F/NqL+bUPc2svL86vb89OBQAA8DPfrsQBAAD4GREHAADgQUQcAACABxFxAAAAHkTEAQAAeBARBwAA4EFEHAAAgAcRcQAAAB70/9yKDbbSBibTAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 722.941x480 with 1 Axes>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "source": [
    "# ep.plot_rgb(\n",
    "#     arr_st,\n",
    "#     rgb=(3, 2, 1),\n",
    "#     stretch=True,\n",
    "#     str_clip=0.02,\n",
    "#     figsize=(12, 16),\n",
    "#     # title=\"RGB Composite Image with Stretch Applied\",\n",
    "# )\n",
    "\n",
    "# plt.show()\n",
    "tt = arr_st[2]\n",
    "print(tt.shape)\n",
    "tt\n",
    "plt.imshow(tt)\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "(2122, 3196)\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "source": [
    "def applyPCA(X, numComponents=75):\n",
    "    newX = np.reshape(X, (-1, X.shape[2]))\n",
    "    pca = PCA(n_components=numComponents, whiten=True)\n",
    "    newX = pca.fit_transform(newX)\n",
    "    newX = np.reshape(newX, (X.shape[0],X.shape[1], numComponents))\n",
    "    return newX, pca"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "source": [
    "def padWithZeros(X, margin=2):\n",
    "    newX = np.zeros((X.shape[0] + 2 * margin, X.shape[1] + 2* margin, X.shape[2]))\n",
    "    x_offset = margin\n",
    "    y_offset = margin\n",
    "    newX[x_offset:X.shape[0] + x_offset, y_offset:X.shape[1] + y_offset, :] = X\n",
    "    return newX"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "source": [
    "def createImageCubes(X, y, windowSize=5, removeZeroLabels = False):\n",
    "    margin = int((windowSize - 1) / 2)\n",
    "    zeroPaddedX = padWithZeros(X, margin=margin)\n",
    "    # split patches\n",
    "    patchesData = np.zeros((X.shape[0] * X.shape[1], windowSize, windowSize, X.shape[2]))\n",
    "    patchesLabels = np.zeros((X.shape[0] * X.shape[1]))\n",
    "    patchIndex = 0\n",
    "    for r in range(margin, zeroPaddedX.shape[0] - margin):\n",
    "        for c in range(margin, zeroPaddedX.shape[1] - margin):\n",
    "            patch = zeroPaddedX[r - margin:r + margin + 1, c - margin:c + margin + 1]   \n",
    "            patchesData[patchIndex, :, :, :] = patch\n",
    "            patchesLabels[patchIndex] = y[r-margin, c-margin]\n",
    "            patchIndex = patchIndex + 1\n",
    "    if removeZeroLabels:\n",
    "        patchesData = patchesData[patchesLabels>0,:,:,:]\n",
    "        patchesLabels = patchesLabels[patchesLabels>0]\n",
    "        patchesLabels -= 1\n",
    "    return patchesData, patchesLabels"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "source": [
    "def splitTrainTestSet(X, y, testRatio, randomState=42):\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=testRatio, random_state=randomState)\n",
    "    return X_train, X_test, y_train, y_test"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "source": [
    "from sklearn.decomposition import PCA\n",
    "from sklearn.model_selection import train_test_split\n",
    "import tensorflow as tf\n",
    "dataset = 'SB'\n",
    "test_size = 0.30\n",
    "windowSize = 5\n",
    "MODEL_NAME = 'Sundarbans'\n",
    "path = '/content/drive/MyDrive/Sundarbans/'\n",
    "\n",
    "X_data = np.moveaxis(arr_st, 0, -1)\n",
    "#y_data = loadmat('Sundarbands_gt.mat')['gt']\n",
    "\n",
    "# Apply PCA\n",
    "K = 3\n",
    "X,pca = applyPCA(X_data,numComponents=K)\n",
    "\n",
    "print(f'Data After PCA: {X.shape}')\n",
    "\n",
    "# Create 3D Patches\n",
    "X, y = createImageCubes(X, y_data, windowSize=windowSize)\n",
    "print(f'Patch size: {X.shape}')\n",
    "\n",
    "print(f'test_size: {test_size}')\n",
    "\n",
    "# Split train and test\n",
    "#X_train, X_test, y_train, y_test = splitTrainTestSet(X, y, testRatio = test_size)\n",
    "X_train = X[:int(len(X)*.8)]\n",
    "y_train = y[:int(len(X)*.8)]\n",
    "X_test = X[int(len(X)*.8):]\n",
    "y_test = y[int(len(X)*.8):]\n",
    "print(\"after train test\")\n",
    "X_train = X_train.reshape(-1, windowSize, windowSize, K, 1)\n",
    "X_test = X_test.reshape(-1, windowSize, windowSize, K, 1)\n",
    "print(\"after X\")\n",
    "\n",
    "# One Hot Encoding\n",
    "y_train = tf.keras.utils.to_categorical(y_train)\n",
    "y_test = tf.keras.utils.to_categorical(y_test)\n",
    "\n",
    "print(f'Train: {X_train.shape}\\nTest: {X_test.shape}\\nTrain Labels: {y_train.shape}\\nTest Labels: {y_test.shape}')\n"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Data After PCA: (2122, 3196, 3)\n",
      "Patch size: (6781912, 5, 5, 3)\n",
      "test_size: 0.3\n",
      "after train test\n",
      "after X\n",
      "Train: (5425529, 5, 5, 3, 1)\n",
      "Test: (1356383, 5, 5, 3, 1)\n",
      "Train Labels: (5425529, 5)\n",
      "Test Labels: (1356383, 5)\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [
    "np"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "<module 'numpy' from '/usr/local/lib/python3.8/dist-packages/numpy/__init__.py'>"
      ]
     },
     "metadata": {},
     "execution_count": 14
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "source": [
    "from tensorflow import keras\n",
    "from keras.layers import Conv2D, Conv3D, Flatten, Dense, Reshape, BatchNormalization\n",
    "from keras.layers import Dropout, Input\n",
    "from keras.models import Model\n",
    "#from keras.optimizers import Adam\n",
    "#from keras.callbacks import ModelCheckpoint\n",
    "#from keras.utils import np_utils\n",
    "#from tensorflow.keras.layers import \n",
    "# windowSize = 15 \n",
    "# K = 1\n",
    "S = windowSize\n",
    "L = K\n",
    "output_units = y_train.shape[1]\n",
    "\n",
    "## input layer\n",
    "input_layer = Input((S, S, L, 1))\n",
    "print(input_layer)\n",
    "## convolutional layers\n",
    "conv_layer1 = Conv3D(filters=16, kernel_size=(2, 2, 1), activation='relu')(input_layer)\n",
    "conv_layer2 = Conv3D(filters=32, kernel_size=(2, 2, 1), activation='relu')(conv_layer1)\n",
    "conv2d_shape = conv_layer2.shape\n",
    "conv_layer3 = Reshape((conv2d_shape[1], conv2d_shape[2], conv2d_shape[3]*conv2d_shape[4]))(conv_layer2)\n",
    "conv_layer4 = Conv2D(filters=64, kernel_size=(2,2), activation='relu')(conv_layer3)\n",
    "\n",
    "flatten_layer = Flatten()(conv_layer4)\n",
    "\n",
    "## fully connected layers\n",
    "dense_layer1 = Dense(128, activation='relu')(flatten_layer)\n",
    "dense_layer1 = Dropout(0.4)(dense_layer1)\n",
    "dense_layer2 = Dense(64, activation='relu')(dense_layer1)\n",
    "dense_layer2 = Dropout(0.4)(dense_layer2)\n",
    "dense_layer3 = Dense(20, activation='relu')(dense_layer2)\n",
    "dense_layer3 = Dropout(0.4)(dense_layer3)\n",
    "output_layer = Dense(units=output_units, activation='softmax')(dense_layer3)\n",
    "\n",
    "import tensorflow as tf\n",
    "# tf.config.gpu.set_per_process_memory_fraction(0.75)\n",
    "# tf.config.gpu.set_per_process_memory_growth(True)\n",
    "\n",
    "# define the model with input layer and output layer\n",
    "model = Model(name = dataset+'_Model' , inputs=input_layer, outputs=output_layer)\n",
    "model.summary()"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "KerasTensor(type_spec=TensorSpec(shape=(None, 5, 5, 3, 1), dtype=tf.float32, name='input_1'), name='input_1', description=\"created by layer 'input_1'\")\n",
      "Model: \"SB_Model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, 5, 5, 3, 1)]      0         \n",
      "                                                                 \n",
      " conv3d (Conv3D)             (None, 4, 4, 3, 16)       80        \n",
      "                                                                 \n",
      " conv3d_1 (Conv3D)           (None, 3, 3, 3, 32)       2080      \n",
      "                                                                 \n",
      " reshape (Reshape)           (None, 3, 3, 96)          0         \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 2, 2, 64)          24640     \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 256)               0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 128)               32896     \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 128)               0         \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 64)                8256      \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 64)                0         \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 20)                1300      \n",
      "                                                                 \n",
      " dropout_2 (Dropout)         (None, 20)                0         \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 5)                 105       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 69,357\n",
      "Trainable params: 69,357\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "source": [
    "def keras_model_memory_usage_in_bytes(model, *, batch_size: int):\n",
    "    \"\"\"\n",
    "    Return the estimated memory usage of a given Keras model in bytes.\n",
    "    This includes the model weights and layers, but excludes the dataset.\n",
    "\n",
    "    The model shapes are multipled by the batch size, but the weights are not.\n",
    "\n",
    "    Args:\n",
    "        model: A Keras model.\n",
    "        batch_size: The batch size you intend to run the model with. If you\n",
    "            have already specified the batch size in the model itself, then\n",
    "            pass `1` as the argument here.\n",
    "    Returns:\n",
    "        An estimate of the Keras model's memory usage in bytes.\n",
    "\n",
    "    \"\"\"\n",
    "    default_dtype = tf.keras.backend.floatx()\n",
    "    shapes_mem_count = 0\n",
    "    internal_model_mem_count = 0\n",
    "    for layer in model.layers:\n",
    "        if isinstance(layer, tf.keras.Model):\n",
    "            internal_model_mem_count += keras_model_memory_usage_in_bytes(\n",
    "                layer, batch_size=batch_size\n",
    "            )\n",
    "        single_layer_mem = tf.as_dtype(layer.dtype or default_dtype).size\n",
    "        out_shape = layer.output_shape\n",
    "        if isinstance(out_shape, list):\n",
    "            out_shape = out_shape[0]\n",
    "        for s in out_shape:\n",
    "            if s is None:\n",
    "                continue\n",
    "            single_layer_mem *= s\n",
    "        shapes_mem_count += single_layer_mem\n",
    "\n",
    "    trainable_count = sum(\n",
    "        [tf.keras.backend.count_params(p) for p in model.trainable_weights]\n",
    "    )\n",
    "    non_trainable_count = sum(\n",
    "        [tf.keras.backend.count_params(p) for p in model.non_trainable_weights]\n",
    "    )\n",
    "\n",
    "    total_memory = (\n",
    "        batch_size * shapes_mem_count\n",
    "        + internal_model_mem_count\n",
    "        + trainable_count\n",
    "        + non_trainable_count\n",
    "    )\n",
    "    return total_memory"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "source": [
    "ufdsa = keras_model_memory_usage_in_bytes(model, batch_size=128)\n",
    "print(ufdsa)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "1867501\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "source": [
    "#!export CUDA_VISIBLE_DEVICES=1\n",
    "# gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=0.333) \n",
    "# sess = tf.Session(config=tf.ConfigProto(log_device_placement=True, gpu_options=gpu_options))\n",
    "tf.config.experimental.get_memory_info('GPU:0')\n",
    "from datetime import datetime\n",
    "from keras.callbacks import TensorBoard\n",
    "from keras.callbacks import ModelCheckpoint, EarlyStopping\n",
    "# Compile\n",
    "model.compile(optimizer = 'adam', loss = 'categorical_crossentropy',\n",
    "              metrics = ['accuracy'])\n",
    "path=\"aspiranture\"\n",
    "# Callbacks\n",
    "logdir = path+\"logs/\" +model.name+'_'+datetime.now().strftime(\"%d:%m:%Y-%H:%M:%S\")\n",
    "\n",
    "tensorboard_callback = TensorBoard(log_dir=logdir)\n",
    "\n",
    "es = EarlyStopping(monitor = 'val_loss',\n",
    "                   min_delta = 0,\n",
    "                   patience = 1,\n",
    "                   verbose = 1,\n",
    "                   restore_best_weights = True)\n",
    "\n",
    "checkpoint = ModelCheckpoint(filepath = 'Pavia_University_Model.h5', \n",
    "                             monitor = 'val_loss', \n",
    "                             mode ='min', \n",
    "                             save_best_only = True,\n",
    "                             verbose = 1)\n",
    "# Fit\n",
    "history = model.fit(x=X_train, y=y_train, \n",
    "                    batch_size=64, epochs=6, \n",
    "                    validation_data=(X_test, y_test), callbacks = [tensorboard_callback, checkpoint])\n"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Epoch 1/6\n",
      "84774/84774 [==============================] - ETA: 0s - loss: 0.1608 - accuracy: 0.9657\n",
      "Epoch 1: val_loss improved from inf to 0.28439, saving model to Pavia_University_Model.h5\n",
      "84774/84774 [==============================] - 744s 9ms/step - loss: 0.1608 - accuracy: 0.9657 - val_loss: 0.2844 - val_accuracy: 0.9398\n",
      "Epoch 2/6\n",
      "84773/84774 [============================>.] - ETA: 0s - loss: 0.1554 - accuracy: 0.9668\n",
      "Epoch 2: val_loss did not improve from 0.28439\n",
      "84774/84774 [==============================] - 752s 9ms/step - loss: 0.1554 - accuracy: 0.9668 - val_loss: 0.2959 - val_accuracy: 0.9389\n",
      "Epoch 3/6\n",
      "84770/84774 [============================>.] - ETA: 0s - loss: 0.1550 - accuracy: 0.9669\n",
      "Epoch 3: val_loss did not improve from 0.28439\n",
      "84774/84774 [==============================] - 732s 9ms/step - loss: 0.1550 - accuracy: 0.9669 - val_loss: 0.2904 - val_accuracy: 0.9424\n",
      "Epoch 4/6\n",
      "84768/84774 [============================>.] - ETA: 0s - loss: 0.1558 - accuracy: 0.9667\n",
      "Epoch 4: val_loss did not improve from 0.28439\n",
      "84774/84774 [==============================] - 746s 9ms/step - loss: 0.1558 - accuracy: 0.9667 - val_loss: 0.2942 - val_accuracy: 0.9415\n",
      "Epoch 5/6\n",
      "84773/84774 [============================>.] - ETA: 0s - loss: 0.1561 - accuracy: 0.9666\n",
      "Epoch 5: val_loss improved from 0.28439 to 0.28306, saving model to Pavia_University_Model.h5\n",
      "84774/84774 [==============================] - 748s 9ms/step - loss: 0.1561 - accuracy: 0.9666 - val_loss: 0.2831 - val_accuracy: 0.9417\n",
      "Epoch 6/6\n",
      "84768/84774 [============================>.] - ETA: 0s - loss: 0.1560 - accuracy: 0.9667\n",
      "Epoch 6: val_loss improved from 0.28306 to 0.27953, saving model to Pavia_University_Model.h5\n",
      "84774/84774 [==============================] - 736s 9ms/step - loss: 0.1560 - accuracy: 0.9667 - val_loss: 0.2795 - val_accuracy: 0.9421\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "source": [
    "import pandas as pd\n",
    "history11 = pd.DataFrame(history.history)\n",
    "\n",
    "fig, axs = plt.subplots(2)\n",
    "axs[0].plot(range(len(history11['accuracy'].values.tolist())), history11['accuracy'].values.tolist(), label = 'Train_Accuracy')\n",
    "# plt.plot(range(len(history['loss'].values.tolist())), history['loss'].values.tolist(), label = 'Train_Loss')\n",
    "axs[0].plot(range(len(history11['val_accuracy'].values.tolist())), history11['val_accuracy'].values.tolist(), label = 'Test_Accuracy')\n",
    "# plt.plot(range(len(history['val_loss'].values.tolist())), history['val_loss'].values.tolist(), label = 'Test_Loss')\n",
    "# axs[0].xlabel('Epochs')\n",
    "# axs[0].ylabel('Value')\n",
    "# axs[0].legend()\n",
    "# axs.show()"
   ],
   "outputs": [
    {
     "output_type": "error",
     "ename": "AttributeError",
     "evalue": "'AxesSubplot' object has no attribute 'show'",
     "traceback": [
      "\u001B[0;31m---------------------------------------------------------------------------\u001B[0m",
      "\u001B[0;31mAttributeError\u001B[0m                            Traceback (most recent call last)",
      "Input \u001B[0;32mIn [58]\u001B[0m, in \u001B[0;36m<module>\u001B[0;34m\u001B[0m\n\u001B[1;32m      7\u001B[0m axs[\u001B[38;5;241m0\u001B[39m]\u001B[38;5;241m.\u001B[39mplot(\u001B[38;5;28mrange\u001B[39m(\u001B[38;5;28mlen\u001B[39m(history11[\u001B[38;5;124m'\u001B[39m\u001B[38;5;124mval_accuracy\u001B[39m\u001B[38;5;124m'\u001B[39m]\u001B[38;5;241m.\u001B[39mvalues\u001B[38;5;241m.\u001B[39mtolist())), history11[\u001B[38;5;124m'\u001B[39m\u001B[38;5;124mval_accuracy\u001B[39m\u001B[38;5;124m'\u001B[39m]\u001B[38;5;241m.\u001B[39mvalues\u001B[38;5;241m.\u001B[39mtolist(), label \u001B[38;5;241m=\u001B[39m \u001B[38;5;124m'\u001B[39m\u001B[38;5;124mTest_Accuracy\u001B[39m\u001B[38;5;124m'\u001B[39m)\n\u001B[1;32m      8\u001B[0m \u001B[38;5;66;03m# plt.plot(range(len(history['val_loss'].values.tolist())), history['val_loss'].values.tolist(), label = 'Test_Loss')\u001B[39;00m\n\u001B[1;32m      9\u001B[0m \u001B[38;5;66;03m# axs[0].xlabel('Epochs')\u001B[39;00m\n\u001B[1;32m     10\u001B[0m \u001B[38;5;66;03m# axs[0].ylabel('Value')\u001B[39;00m\n\u001B[1;32m     11\u001B[0m \u001B[38;5;66;03m# axs[0].legend()\u001B[39;00m\n\u001B[1;32m     12\u001B[0m \u001B[38;5;66;03m# axs.show()\u001B[39;00m\n\u001B[0;32m---> 13\u001B[0m \u001B[43maxs\u001B[49m\u001B[43m[\u001B[49m\u001B[38;5;241;43m0\u001B[39;49m\u001B[43m]\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mshow\u001B[49m()\n",
      "\u001B[0;31mAttributeError\u001B[0m: 'AxesSubplot' object has no attribute 'show'"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "source": [
    "!pip install seaborn\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "pred = model.predict(X_test, batch_size=1204*6, verbose=1)\n",
    "\n",
    "plt.figure(figsize = (10,7))\n",
    "\n",
    "classes = [f'Class-{i}' for i in range(1, 6)]\n",
    "\n",
    "mat = confusion_matrix(np.argmax(y_test, 1),\n",
    "                            np.argmax(pred, 1))\n",
    "\n",
    "df_cm = pd.DataFrame(mat, index = classes, columns = classes)\n",
    "\n",
    "sns.heatmap(df_cm, annot=True, fmt='d')\n",
    "\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Requirement already satisfied: seaborn in /usr/local/lib/python3.8/dist-packages (0.11.2)\n",
      "Requirement already satisfied: pandas>=0.23 in /usr/local/lib/python3.8/dist-packages (from seaborn) (1.4.0)\n",
      "Requirement already satisfied: numpy>=1.15 in /usr/local/lib/python3.8/dist-packages (from seaborn) (1.22.1)\n",
      "Requirement already satisfied: matplotlib>=2.2 in /usr/local/lib/python3.8/dist-packages (from seaborn) (3.5.1)\n",
      "Requirement already satisfied: scipy>=1.0 in /usr/local/lib/python3.8/dist-packages (from seaborn) (1.8.0)\n",
      "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.23->seaborn) (2021.3)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.23->seaborn) (2.8.2)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.2->seaborn) (3.0.7)\n",
      "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.2->seaborn) (21.3)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.2->seaborn) (9.0.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.2->seaborn) (1.3.2)\n",
      "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.2->seaborn) (0.11.0)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib>=2.2->seaborn) (4.29.1)\n",
      "Requirement already satisfied: six>=1.5 in /usr/lib/python3/dist-packages (from python-dateutil>=2.8.1->pandas>=0.23->seaborn) (1.14.0)\n",
      "\u001B[33mWARNING: You are using pip version 20.2.4; however, version 22.0.3 is available.\n",
      "You should consider upgrading via the '/usr/bin/python3 -m pip install --upgrade pip' command.\u001B[0m\n",
      "188/188 [==============================] - 1s 5ms/step\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x700 with 2 Axes>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "source": [
    "!gdown --id \"1evrBQtvdFV52dDkLVIr31pVr8aRfNOYz\"\n",
    "!unzip ./datasett.zip"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Downloading...\n",
      "From: https://drive.google.com/uc?id=1evrBQtvdFV52dDkLVIr31pVr8aRfNOYz\n",
      "To: /tf/datasett.zip\n",
      "100%|██████████████████████████████████████| 33.2k/33.2k [00:00<00:00, 2.37MB/s]\n",
      "Archive:  ./datasett.zip\n",
      "replace datasett/test.png? [y]es, [n]o, [A]ll, [N]one, [r]ename: ^C\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "source": [
    "files_test = glob(\"datasett/*.png\")\n",
    "files_test.sort()\n",
    "files_test"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "['datasett/test.png']"
      ]
     },
     "metadata": {},
     "execution_count": 21
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "source": [
    "# lt = []\n",
    "# for i in files_test:\n",
    "#   with rio.open(i, 'r') as f:\n",
    "#     lt.append(f.read(1))\n",
    "lt = []\n",
    "for i in files_test:\n",
    "  im = plt.imread(i)\n",
    "  lt.append(im[:, :, 0])\n",
    "  lt.append(im[:, :, 1])\n",
    "  lt.append(im[:, :, 2])"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "source": [
    "arr_tst = np.stack(lt)"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "source": [
    "plt.imshow(lt[1])\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "source": [
    "X_data_validate = np.moveaxis(arr_tst, 0, -1)"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "source": [
    "X_validate,pca = applyPCA(X_data_validate,numComponents=K)"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "source": [
    "X_validate11, y = createImageCubes(X_validate, y_data, windowSize=windowSize)"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "source": [
    "X_validate111 = X_validate11.reshape(-1, windowSize, windowSize, K, 1)"
   ],
   "outputs": [],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "source": [
    "pred_validate = model.predict(X_validate111, batch_size=128, verbose=1)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "23487/23487 [==============================] - 61s 3ms/step\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "source": [
    "print(pred_validate.shape)\n",
    "bb = np.argmax(pred_validate, 1)\n",
    "print(np.count_nonzero(bb == 4))\n",
    "bbb = bb.reshape(1478,2034)\n",
    "# pred_vall = np.argmax(pred_validate,1)\n",
    "#print(pred_vall)\n",
    "#preed = pred_validate/pred_validate.max()\n",
    "#fig = plt.figure(figsize=(6, 3.2))\n",
    "\n",
    "#ax = fig.add_subplot(111)\n",
    "#hh = np.array(pred_validate)\n",
    "plt.matshow(bbb, cmap='brg')\n",
    "plt.show()\n",
    "#ww = lt[0]\n",
    "#ww.shape\n",
    "#preddd = pred_validate.reshape(1478, 2034)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "(3006252, 5)\n",
      "39\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 660.568x480 with 1 Axes>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [
    "#print(pred)\n",
    "pred1 = np.argmax(pred, axis=1)\n",
    "print(pred1)\n",
    "print(np.max(pred1))\n",
    "print(np.size(pred1))\n",
    "print(np.size(X_test))\n",
    "\n",
    "#plt.matshow(pred1)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "[1 1 1 ... 0 0 0]\n",
      "4\n",
      "1356383\n",
      "109867023\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [
    "from sklearn.metrics import classification_report\n",
    "y_pred_test = np.argmax(pred, axis=1)\n",
    "\n",
    "classification = classification_report(np.argmax(y_test, axis=1), y_pred_test)\n",
    "print(classification)"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.55      0.29      0.38     69713\n",
      "           1       0.95      0.99      0.97    876052\n",
      "           2       0.95      0.98      0.96     71533\n",
      "           3       0.96      0.94      0.95    214563\n",
      "           4       0.97      0.98      0.98    124522\n",
      "\n",
      "    accuracy                           0.94   1356383\n",
      "   macro avg       0.88      0.84      0.85   1356383\n",
      "weighted avg       0.93      0.94      0.94   1356383\n",
      "\n"
     ]
    }
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.8.10",
   "mimetype": "text/x-python",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "pygments_lexer": "ipython3",
   "nbconvert_exporter": "python",
   "file_extension": ".py"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}