Amazon Kaggle Project.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5"
   },
   "outputs": [],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load\n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "\n",
    "# Input data files are available in the read-only \"../input/\" directory\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n",
    "\n",
    "import os\n",
    "\n",
    "\n",
    "# You can write up to 5GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n",
    "# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "_cell_guid": "79c7e3d0-c299-4dcb-8224-4455121ee9b0",
    "_uuid": "d629ff2d2480ee46fbb7e2d37f6b5fab8052498a"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from tensorflow.keras.models import Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train_0</td>\n",
       "      <td>haze primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train_1</td>\n",
       "      <td>agriculture clear primary water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>train_2</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>train_3</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>train_4</td>\n",
       "      <td>agriculture clear habitation primary road</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  image_name                                       tags\n",
       "0    train_0                               haze primary\n",
       "1    train_1            agriculture clear primary water\n",
       "2    train_2                              clear primary\n",
       "3    train_3                              clear primary\n",
       "4    train_4  agriculture clear habitation primary road"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_classes_df = pd.read_csv('../input/planets-dataset/planet/planet/train_classes.csv')\n",
    "train_classes_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "40479\r\n"
     ]
    }
   ],
   "source": [
    "!ls ../input/planets-dataset/planet/planet/train-jpg | wc -l"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test_0</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test_1</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test_2</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test_3</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test_4</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  image_name                                  tags\n",
       "0     test_0  primary clear agriculture road water\n",
       "1     test_1  primary clear agriculture road water\n",
       "2     test_2  primary clear agriculture road water\n",
       "3     test_3  primary clear agriculture road water\n",
       "4     test_4  primary clear agriculture road water"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_submission_df = pd.read_csv('../input/planets-dataset/planet/planet/sample_submission.csv')\n",
    "sample_submission_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "test1 = !ls ../input/planets-dataset/planet/planet/test-jpg | wc -l\n",
    "test2 = !ls ../input/planets-dataset/test-jpg-additional/test-jpg-additional | wc -l\n",
    "assert sample_submission_df.shape[0] == float(test1[0])+float(test2[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['20522']"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from skimage import io\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_number =10\n",
    "img = io.imread('../input/planets-dataset/planet/planet/train-jpg/train_{}.jpg'.format(image_number))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(256, 256, 3)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fd9cfb9ea10>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>train_10</td>\n",
       "      <td>agriculture clear primary slash_burn water</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   image_name                                        tags\n",
       "10   train_10  agriculture clear primary slash_burn water"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_classes_df[train_classes_df['image_name'] == 'train_10']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train_0</td>\n",
       "      <td>haze primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train_1</td>\n",
       "      <td>agriculture clear primary water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>train_2</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>train_3</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>train_4</td>\n",
       "      <td>agriculture clear habitation primary road</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  image_name                                       tags\n",
       "0    train_0                               haze primary\n",
       "1    train_1            agriculture clear primary water\n",
       "2    train_2                              clear primary\n",
       "3    train_3                              clear primary\n",
       "4    train_4  agriculture clear habitation primary road"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_classes_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['conventional_mine', 'habitation', 'partly_cloudy', 'clear', 'slash_burn', 'cloudy', 'haze', 'blow_down', 'road', 'agriculture', 'selective_logging', 'bare_ground', 'cultivation', 'primary', 'artisinal_mine', 'water', 'blooming']\n"
     ]
    }
   ],
   "source": [
    "unique_labels = set()\n",
    "def append_labels(tags):\n",
    "    for tag in tags.split():\n",
    "        unique_labels.add(tag)\n",
    "\n",
    "train_classes = train_classes_df.copy()\n",
    "train_classes['tags'].apply(append_labels)\n",
    "unique_labels = list(unique_labels)\n",
    "print(unique_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(unique_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert len(train_classes['image_name'].unique()) == train_classes.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "      <th>conventional_mine</th>\n",
       "      <th>habitation</th>\n",
       "      <th>partly_cloudy</th>\n",
       "      <th>clear</th>\n",
       "      <th>slash_burn</th>\n",
       "      <th>cloudy</th>\n",
       "      <th>haze</th>\n",
       "      <th>blow_down</th>\n",
       "      <th>road</th>\n",
       "      <th>agriculture</th>\n",
       "      <th>selective_logging</th>\n",
       "      <th>bare_ground</th>\n",
       "      <th>cultivation</th>\n",
       "      <th>primary</th>\n",
       "      <th>artisinal_mine</th>\n",
       "      <th>water</th>\n",
       "      <th>blooming</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train_0.jpg</td>\n",
       "      <td>haze primary</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train_1.jpg</td>\n",
       "      <td>agriculture clear primary water</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>train_2.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>train_3.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>train_4.jpg</td>\n",
       "      <td>agriculture clear habitation primary road</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    image_name                                       tags  conventional_mine  \\\n",
       "0  train_0.jpg                               haze primary                  0   \n",
       "1  train_1.jpg            agriculture clear primary water                  0   \n",
       "2  train_2.jpg                              clear primary                  0   \n",
       "3  train_3.jpg                              clear primary                  0   \n",
       "4  train_4.jpg  agriculture clear habitation primary road                  0   \n",
       "\n",
       "   habitation  partly_cloudy  clear  slash_burn  cloudy  haze  blow_down  \\\n",
       "0           0              0      0           0       0     1          0   \n",
       "1           0              0      1           0       0     0          0   \n",
       "2           0              0      1           0       0     0          0   \n",
       "3           0              0      1           0       0     0          0   \n",
       "4           1              0      1           0       0     0          0   \n",
       "\n",
       "   road  agriculture  selective_logging  bare_ground  cultivation  primary  \\\n",
       "0     0            0                  0            0            0        1   \n",
       "1     0            1                  0            0            0        1   \n",
       "2     0            0                  0            0            0        1   \n",
       "3     0            0                  0            0            0        1   \n",
       "4     1            1                  0            0            0        1   \n",
       "\n",
       "   artisinal_mine  water  blooming  \n",
       "0               0      0         0  \n",
       "1               0      1         0  \n",
       "2               0      0         0  \n",
       "3               0      0         0  \n",
       "4               0      0         0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's do one hot encoding (vectorize) the labels in 'train_classes'\n",
    "for tag in unique_labels:\n",
    "    train_classes[tag] = train_classes['tags'].apply(lambda x: 1 if tag in x.split() else 0)\n",
    "    \n",
    "# adding '.jpg' extension to 'image_name'\n",
    "train_classes['image_name'] = train_classes['image_name'].apply(lambda x: '{}.jpg'.format(x)) \n",
    "train_classes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 40479 validated image filenames.\n"
     ]
    }
   ],
   "source": [
    "y_col = list(train_classes.columns[2:]) # storing the tags column names as a variable\n",
    "\n",
    "# initializing an image generator with some data augumentation\n",
    "image_gen = tf.keras.preprocessing.image.ImageDataGenerator(rescale=1/255)\n",
    "\n",
    "# loading images from dataframe\n",
    "X = image_gen.flow_from_dataframe(dataframe=train_classes, \\\n",
    "        directory='/kaggle/input/planets-dataset/planet/planet/train-jpg/', x_col='image_name', y_col=y_col, \\\n",
    "       target_size=(64, 64), class_mode='raw', seed=1, batch_size=64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "each image's shape is (64, 64, 3)\n",
      "each label's shape is (17,)\n",
      "we have 633 batches\n",
      "each batch has 64 images/labels\n",
      "40479/128 is 316.24, so the last batch will have 64 images/labels\n"
     ]
    }
   ],
   "source": [
    "# X is an iterable, It contains 317 batches, each batch contains 128 images and labels because \n",
    "#40479 / 128 is 316 remainder 31 each image is of shape (128, 128, 3), each label is of shape (17, )\n",
    "\n",
    "# let's abitrarily view an image\n",
    "x54 = X[0][0][54] # first batch, images, 109th image\n",
    "y54 = X[0][1][54] # first batch, labels, 109th label\n",
    "print(\"each image's shape is {}\".format(x54.shape))\n",
    "print(\"each label's shape is {}\".format(y54.shape))\n",
    "print('we have {} batches'.format(len(X)))\n",
    "print('each batch has {} images/labels'.format(X[0][0].shape[0]))\n",
    "print('40479/128 is {:.2F}, so the last batch will have {} images/labels'.format(40479/128, X[316][0].shape[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Flatten, BatchNormalization, Conv2D\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "from keras.optimizers import SGD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fbeta(ytrue , ypred, beta=2, epsilon=1e-4):\n",
    "    beta_squarred = beta**2\n",
    "\n",
    "    ytrue = tf.cast(ytrue, tf.float32)\n",
    "    ypred = tf.cast(tf.greater(tf.cast(ypred, tf.float32), tf.constant(0.5)), tf.float32)\n",
    "        \n",
    "    tp = tf.reduce_sum(ytrue * ypred, axis=1)\n",
    "    fp = tf.reduce_sum(ypred, axis=1) - tp\n",
    "    fn = tf.reduce_sum(ytrue, axis=1) - tp\n",
    "    \n",
    "    precision = tp/(tp+fp+epsilon)\n",
    "    recall = tp/(tp+fn+epsilon)\n",
    "    \n",
    "    fb = (1+beta_squarred)*precision*recall / (beta_squarred*precision + recall + epsilon)\n",
    "    return fb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def multi_label_acc(ytrue , ypred, epsilon=1e-4):\n",
    "    \n",
    "    ytrue = tf.cast(ytrue, tf.float32)\n",
    "    ypred = tf.cast(tf.greater(tf.cast(ypred, tf.float32), tf.constant(0.5)), tf.float32)\n",
    "    \n",
    "    tp = tf.reduce_sum(ytrue * ypred, axis=1)\n",
    "    fp = tf.reduce_sum(ypred, axis=1) - tp\n",
    "    fn = tf.reduce_sum(ytrue, axis=1) - tp\n",
    "    \n",
    "    ytrue = tf.cast(ytrue, tf.bool)\n",
    "    ypred = tf.cast(ypred, tf.bool)\n",
    "    \n",
    "    tn = tf.reduce_sum(tf.cast(tf.logical_not(ytrue), tf.float32) * tf.cast(tf.logical_not(ypred), tf.float32),\\\n",
    "                       axis=1)\n",
    "    \n",
    "    return (tp+tn)/(tp+tn+fp+fn+epsilon)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model():\n",
    "    in_shape=(64, 64, 3)\n",
    "    out_shape=17\n",
    "\n",
    "    # load mannequin\n",
    "\n",
    "    mannequin = VGG16(include_top=False, input_shape=in_shape)\n",
    "\n",
    "    # mark loaded layers as not trainable\n",
    "\n",
    "    for layer in mannequin.layers:\n",
    "        layer.trainable = False\n",
    "\n",
    "    # enable final vgg block to be trainable\n",
    "\n",
    "    mannequin.get_layer(\"block5_conv1\").trainable = True\n",
    "\n",
    "    mannequin.get_layer(\"block5_conv2\").trainable = True\n",
    "\n",
    "    mannequin.get_layer(\"block5_conv3\").trainable = True\n",
    "\n",
    "    mannequin.get_layer(\"block5_pool\").trainable = True\n",
    "\n",
    "    # add new classifier layers\n",
    "\n",
    "    flat1 = Flatten()(mannequin.layers[-1].output)\n",
    "\n",
    "    class1 = Dense(128, activation=\"relu\", kernel_initializer=\"he_uniform\")(flat1)\n",
    "\n",
    "    output = Dense(out_shape, activation=\"sigmoid\")(class1)\n",
    "\n",
    "    # outline new mannequin\n",
    "\n",
    "    mannequin = Model(inputs=mannequin.inputs, outputs=output)\n",
    "\n",
    "    # compile mannequin\n",
    "\n",
    "    choose = SGD(lr=0.01, momentum=0.9)\n",
    "\n",
    "    mannequin.compile(optimizer=choose, loss=\"binary_crossentropy\", metrics=[fbeta, multi_label_acc])\n",
    "    \n",
    "    return mannequin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_best_check_point = ModelCheckpoint(filepath='best_model.hdf5', monitor='val_fbeta',mode='max', save_best_only=True, save_weights_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 32384 validated image filenames.\n",
      "Found 8095 validated image filenames.\n"
     ]
    }
   ],
   "source": [
    "train_image_gen = tf.keras.preprocessing.image.ImageDataGenerator(rescale=1/255,validation_split=0.2)\n",
    "\n",
    "# generating the 80% training image data\n",
    "train_gen = train_image_gen.flow_from_dataframe(dataframe=train_classes, \\\n",
    "        directory='../input/planets-dataset/planet/planet/train-jpg/', x_col='image_name', y_col=y_col, \\\n",
    "       target_size=(64, 64), class_mode='raw', seed=0, batch_size=64, subset='training')\n",
    "\n",
    "# generating the 20% validation image data\n",
    "val_gen = train_image_gen.flow_from_dataframe(dataframe=train_classes, \\\n",
    "        directory='../input/planets-dataset/planet/planet/train-jpg/', x_col='image_name', y_col=y_col, \\\n",
    "       target_size=(64,64), class_mode='raw', seed=0, batch_size=64, subset='validation')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# setting step size for training and validation image data\n",
    "step_train_size = int(np.ceil(train_gen.samples / train_gen.batch_size))\n",
    "step_val_size = int(np.ceil(val_gen.samples / train_gen.batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "633"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "step_train_size+step_val_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
      "58892288/58889256 [==============================] - 1s 0us/step\n",
      "Epoch 1/10\n",
      "506/506 [==============================] - 132s 260ms/step - loss: 0.1724 - fbeta: 0.7657 - multi_label_acc: 0.9321 - val_loss: 0.1509 - val_fbeta: 0.7809 - val_multi_label_acc: 0.9407\n",
      "Epoch 2/10\n",
      "506/506 [==============================] - 107s 212ms/step - loss: 0.1410 - fbeta: 0.8196 - multi_label_acc: 0.9452 - val_loss: 0.1353 - val_fbeta: 0.8256 - val_multi_label_acc: 0.9471\n",
      "Epoch 3/10\n",
      "506/506 [==============================] - 106s 210ms/step - loss: 0.1343 - fbeta: 0.8295 - multi_label_acc: 0.9476 - val_loss: 0.1349 - val_fbeta: 0.8214 - val_multi_label_acc: 0.9472\n",
      "Epoch 4/10\n",
      "506/506 [==============================] - 106s 209ms/step - loss: 0.1297 - fbeta: 0.8374 - multi_label_acc: 0.9494 - val_loss: 0.1292 - val_fbeta: 0.8477 - val_multi_label_acc: 0.9503\n",
      "Epoch 5/10\n",
      "506/506 [==============================] - 106s 209ms/step - loss: 0.1261 - fbeta: 0.8427 - multi_label_acc: 0.9507 - val_loss: 0.1242 - val_fbeta: 0.8456 - val_multi_label_acc: 0.9522\n",
      "Epoch 6/10\n",
      "506/506 [==============================] - 105s 208ms/step - loss: 0.1229 - fbeta: 0.8472 - multi_label_acc: 0.9520 - val_loss: 0.1208 - val_fbeta: 0.8576 - val_multi_label_acc: 0.9528\n",
      "Epoch 7/10\n",
      "506/506 [==============================] - 106s 209ms/step - loss: 0.1202 - fbeta: 0.8517 - multi_label_acc: 0.9530 - val_loss: 0.1213 - val_fbeta: 0.8617 - val_multi_label_acc: 0.9533\n",
      "Epoch 8/10\n",
      "506/506 [==============================] - 106s 210ms/step - loss: 0.1184 - fbeta: 0.8546 - multi_label_acc: 0.9536 - val_loss: 0.1197 - val_fbeta: 0.8529 - val_multi_label_acc: 0.9534\n",
      "Epoch 9/10\n",
      "506/506 [==============================] - 105s 207ms/step - loss: 0.1161 - fbeta: 0.8572 - multi_label_acc: 0.9546 - val_loss: 0.1191 - val_fbeta: 0.8491 - val_multi_label_acc: 0.9534\n",
      "Epoch 10/10\n",
      "506/506 [==============================] - 106s 209ms/step - loss: 0.1155 - fbeta: 0.8579 - multi_label_acc: 0.9550 - val_loss: 0.1187 - val_fbeta: 0.8537 - val_multi_label_acc: 0.9537\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7fd9c4315a90>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.applications.vgg16 import VGG16\n",
    "model1 = build_model() # building a sequential model for training\n",
    "\n",
    "# fitting the model\n",
    "model1.fit(x=train_gen, steps_per_epoch=step_train_size, validation_data=val_gen, validation_steps=step_val_size,epochs=10, callbacks=[save_best_check_point])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# V3 : (vgg16:64,64,3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = \"../input/planets-dataset/planet/planet\"\n",
    "train_data = pd.read_csv(data_path+\"/train_classes.csv\")\n",
    "test_img_path = data_path + \"/test-jpg\"\n",
    "train_img_path = data_path + \"/train-jpg\"\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 40479 entries, 0 to 40478\n",
      "Data columns (total 2 columns):\n",
      " #   Column      Non-Null Count  Dtype \n",
      "---  ------      --------------  ----- \n",
      " 0   image_name  40479 non-null  object\n",
      " 1   tags        40479 non-null  object\n",
      "dtypes: object(2)\n",
      "memory usage: 632.6+ KB\n"
     ]
    }
   ],
   "source": [
    "train_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>40479</td>\n",
       "      <td>40479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>40479</td>\n",
       "      <td>449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>train_38520</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1</td>\n",
       "      <td>13636</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         image_name           tags\n",
       "count         40479          40479\n",
       "unique        40479            449\n",
       "top     train_38520  clear primary\n",
       "freq              1          13636"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "449"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_data['tags'].unique())  # number of categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40479, 2)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.shape # No of Indivials"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['haze',\n",
       " 'primary',\n",
       " 'agriculture',\n",
       " 'clear',\n",
       " 'water',\n",
       " 'habitation',\n",
       " 'road',\n",
       " 'cultivation',\n",
       " 'slash_burn',\n",
       " 'cloudy',\n",
       " 'partly_cloudy',\n",
       " 'conventional_mine',\n",
       " 'bare_ground',\n",
       " 'artisinal_mine',\n",
       " 'blooming',\n",
       " 'selective_logging',\n",
       " 'blow_down']"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_list =[]\n",
    "\n",
    "for tag_str in train_data.tags.values:\n",
    "    labels =tag_str.split(' ')\n",
    "    for label in labels:\n",
    "        if label not in label_list:\n",
    "            label_list.append(label)\n",
    "\n",
    "label_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "      <th>haze</th>\n",
       "      <th>primary</th>\n",
       "      <th>agriculture</th>\n",
       "      <th>clear</th>\n",
       "      <th>water</th>\n",
       "      <th>habitation</th>\n",
       "      <th>road</th>\n",
       "      <th>cultivation</th>\n",
       "      <th>slash_burn</th>\n",
       "      <th>cloudy</th>\n",
       "      <th>partly_cloudy</th>\n",
       "      <th>conventional_mine</th>\n",
       "      <th>bare_ground</th>\n",
       "      <th>artisinal_mine</th>\n",
       "      <th>blooming</th>\n",
       "      <th>selective_logging</th>\n",
       "      <th>blow_down</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train_0</td>\n",
       "      <td>haze primary</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train_1</td>\n",
       "      <td>agriculture clear primary water</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>train_2</td>\n",
       "      <td>clear primary</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>train_3</td>\n",
       "      <td>clear primary</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>train_4</td>\n",
       "      <td>agriculture clear habitation primary road</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  image_name                                       tags  haze  primary  \\\n",
       "0    train_0                               haze primary     1        1   \n",
       "1    train_1            agriculture clear primary water     0        1   \n",
       "2    train_2                              clear primary     0        1   \n",
       "3    train_3                              clear primary     0        1   \n",
       "4    train_4  agriculture clear habitation primary road     0        1   \n",
       "\n",
       "   agriculture  clear  water  habitation  road  cultivation  slash_burn  \\\n",
       "0            0      0      0           0     0            0           0   \n",
       "1            1      1      1           0     0            0           0   \n",
       "2            0      1      0           0     0            0           0   \n",
       "3            0      1      0           0     0            0           0   \n",
       "4            1      1      0           1     1            0           0   \n",
       "\n",
       "   cloudy  partly_cloudy  conventional_mine  bare_ground  artisinal_mine  \\\n",
       "0       0              0                  0            0               0   \n",
       "1       0              0                  0            0               0   \n",
       "2       0              0                  0            0               0   \n",
       "3       0              0                  0            0               0   \n",
       "4       0              0                  0            0               0   \n",
       "\n",
       "   blooming  selective_logging  blow_down  \n",
       "0         0                  0          0  \n",
       "1         0                  0          0  \n",
       "2         0                  0          0  \n",
       "3         0                  0          0  \n",
       "4         0                  0          0  "
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for label in label_list:\n",
    "    train_data[label] =train_data['tags'].apply(lambda x: 1 if label in x.split(' ') else 0)\n",
    "# Display head\n",
    "train_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fd9b8157dd0>"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of label instances\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "train_data[label_list].sum().sort_values().plot.bar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>haze</th>\n",
       "      <th>primary</th>\n",
       "      <th>agriculture</th>\n",
       "      <th>clear</th>\n",
       "      <th>water</th>\n",
       "      <th>habitation</th>\n",
       "      <th>road</th>\n",
       "      <th>cultivation</th>\n",
       "      <th>slash_burn</th>\n",
       "      <th>cloudy</th>\n",
       "      <th>partly_cloudy</th>\n",
       "      <th>conventional_mine</th>\n",
       "      <th>bare_ground</th>\n",
       "      <th>artisinal_mine</th>\n",
       "      <th>blooming</th>\n",
       "      <th>selective_logging</th>\n",
       "      <th>blow_down</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>haze</th>\n",
       "      <td>2697</td>\n",
       "      <td>2670</td>\n",
       "      <td>672</td>\n",
       "      <td>0</td>\n",
       "      <td>613</td>\n",
       "      <td>129</td>\n",
       "      <td>394</td>\n",
       "      <td>202</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>primary</th>\n",
       "      <td>2670</td>\n",
       "      <td>37513</td>\n",
       "      <td>11972</td>\n",
       "      <td>27668</td>\n",
       "      <td>7001</td>\n",
       "      <td>3469</td>\n",
       "      <td>7728</td>\n",
       "      <td>4455</td>\n",
       "      <td>209</td>\n",
       "      <td>0</td>\n",
       "      <td>7175</td>\n",
       "      <td>94</td>\n",
       "      <td>683</td>\n",
       "      <td>324</td>\n",
       "      <td>332</td>\n",
       "      <td>340</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>agriculture</th>\n",
       "      <td>672</td>\n",
       "      <td>11972</td>\n",
       "      <td>12315</td>\n",
       "      <td>9150</td>\n",
       "      <td>2712</td>\n",
       "      <td>2737</td>\n",
       "      <td>6034</td>\n",
       "      <td>3377</td>\n",
       "      <td>119</td>\n",
       "      <td>0</td>\n",
       "      <td>2493</td>\n",
       "      <td>24</td>\n",
       "      <td>225</td>\n",
       "      <td>38</td>\n",
       "      <td>32</td>\n",
       "      <td>65</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>clear</th>\n",
       "      <td>0</td>\n",
       "      <td>27668</td>\n",
       "      <td>9150</td>\n",
       "      <td>28431</td>\n",
       "      <td>5502</td>\n",
       "      <td>3090</td>\n",
       "      <td>6295</td>\n",
       "      <td>3527</td>\n",
       "      <td>173</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>70</td>\n",
       "      <td>747</td>\n",
       "      <td>307</td>\n",
       "      <td>311</td>\n",
       "      <td>308</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>water</th>\n",
       "      <td>613</td>\n",
       "      <td>7001</td>\n",
       "      <td>2712</td>\n",
       "      <td>5502</td>\n",
       "      <td>7411</td>\n",
       "      <td>915</td>\n",
       "      <td>2125</td>\n",
       "      <td>868</td>\n",
       "      <td>24</td>\n",
       "      <td>0</td>\n",
       "      <td>1295</td>\n",
       "      <td>26</td>\n",
       "      <td>206</td>\n",
       "      <td>299</td>\n",
       "      <td>16</td>\n",
       "      <td>49</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>habitation</th>\n",
       "      <td>129</td>\n",
       "      <td>3469</td>\n",
       "      <td>2737</td>\n",
       "      <td>3090</td>\n",
       "      <td>915</td>\n",
       "      <td>3660</td>\n",
       "      <td>2786</td>\n",
       "      <td>895</td>\n",
       "      <td>41</td>\n",
       "      <td>0</td>\n",
       "      <td>441</td>\n",
       "      <td>36</td>\n",
       "      <td>163</td>\n",
       "      <td>29</td>\n",
       "      <td>4</td>\n",
       "      <td>13</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>road</th>\n",
       "      <td>394</td>\n",
       "      <td>7728</td>\n",
       "      <td>6034</td>\n",
       "      <td>6295</td>\n",
       "      <td>2125</td>\n",
       "      <td>2786</td>\n",
       "      <td>8071</td>\n",
       "      <td>1294</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>1382</td>\n",
       "      <td>59</td>\n",
       "      <td>323</td>\n",
       "      <td>110</td>\n",
       "      <td>10</td>\n",
       "      <td>151</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cultivation</th>\n",
       "      <td>202</td>\n",
       "      <td>4455</td>\n",
       "      <td>3377</td>\n",
       "      <td>3527</td>\n",
       "      <td>868</td>\n",
       "      <td>895</td>\n",
       "      <td>1294</td>\n",
       "      <td>4477</td>\n",
       "      <td>126</td>\n",
       "      <td>0</td>\n",
       "      <td>748</td>\n",
       "      <td>4</td>\n",
       "      <td>89</td>\n",
       "      <td>18</td>\n",
       "      <td>35</td>\n",
       "      <td>58</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>slash_burn</th>\n",
       "      <td>3</td>\n",
       "      <td>209</td>\n",
       "      <td>119</td>\n",
       "      <td>173</td>\n",
       "      <td>24</td>\n",
       "      <td>41</td>\n",
       "      <td>36</td>\n",
       "      <td>126</td>\n",
       "      <td>209</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cloudy</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2089</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>partly_cloudy</th>\n",
       "      <td>0</td>\n",
       "      <td>7175</td>\n",
       "      <td>2493</td>\n",
       "      <td>0</td>\n",
       "      <td>1295</td>\n",
       "      <td>441</td>\n",
       "      <td>1382</td>\n",
       "      <td>748</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>7261</td>\n",
       "      <td>28</td>\n",
       "      <td>74</td>\n",
       "      <td>27</td>\n",
       "      <td>17</td>\n",
       "      <td>27</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>conventional_mine</th>\n",
       "      <td>2</td>\n",
       "      <td>94</td>\n",
       "      <td>24</td>\n",
       "      <td>70</td>\n",
       "      <td>26</td>\n",
       "      <td>36</td>\n",
       "      <td>59</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>100</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bare_ground</th>\n",
       "      <td>41</td>\n",
       "      <td>683</td>\n",
       "      <td>225</td>\n",
       "      <td>747</td>\n",
       "      <td>206</td>\n",
       "      <td>163</td>\n",
       "      <td>323</td>\n",
       "      <td>89</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>74</td>\n",
       "      <td>10</td>\n",
       "      <td>862</td>\n",
       "      <td>40</td>\n",
       "      <td>3</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>artisinal_mine</th>\n",
       "      <td>5</td>\n",
       "      <td>324</td>\n",
       "      <td>38</td>\n",
       "      <td>307</td>\n",
       "      <td>299</td>\n",
       "      <td>29</td>\n",
       "      <td>110</td>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>4</td>\n",
       "      <td>40</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>blooming</th>\n",
       "      <td>4</td>\n",
       "      <td>332</td>\n",
       "      <td>32</td>\n",
       "      <td>311</td>\n",
       "      <td>16</td>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>35</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>332</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>selective_logging</th>\n",
       "      <td>5</td>\n",
       "      <td>340</td>\n",
       "      <td>65</td>\n",
       "      <td>308</td>\n",
       "      <td>49</td>\n",
       "      <td>13</td>\n",
       "      <td>151</td>\n",
       "      <td>58</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "      <td>340</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>blow_down</th>\n",
       "      <td>0</td>\n",
       "      <td>98</td>\n",
       "      <td>22</td>\n",
       "      <td>85</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   haze  primary  agriculture  clear  water  habitation  road  \\\n",
       "haze               2697     2670          672      0    613         129   394   \n",
       "primary            2670    37513        11972  27668   7001        3469  7728   \n",
       "agriculture         672    11972        12315   9150   2712        2737  6034   \n",
       "clear                 0    27668         9150  28431   5502        3090  6295   \n",
       "water               613     7001         2712   5502   7411         915  2125   \n",
       "habitation          129     3469         2737   3090    915        3660  2786   \n",
       "road                394     7728         6034   6295   2125        2786  8071   \n",
       "cultivation         202     4455         3377   3527    868         895  1294   \n",
       "slash_burn            3      209          119    173     24          41    36   \n",
       "cloudy                0        0            0      0      0           0     0   \n",
       "partly_cloudy         0     7175         2493      0   1295         441  1382   \n",
       "conventional_mine     2       94           24     70     26          36    59   \n",
       "bare_ground          41      683          225    747    206         163   323   \n",
       "artisinal_mine        5      324           38    307    299          29   110   \n",
       "blooming              4      332           32    311     16           4    10   \n",
       "selective_logging     5      340           65    308     49          13   151   \n",
       "blow_down             0       98           22     85      3           3     2   \n",
       "\n",
       "                   cultivation  slash_burn  cloudy  partly_cloudy  \\\n",
       "haze                       202           3       0              0   \n",
       "primary                   4455         209       0           7175   \n",
       "agriculture               3377         119       0           2493   \n",
       "clear                     3527         173       0              0   \n",
       "water                      868          24       0           1295   \n",
       "habitation                 895          41       0            441   \n",
       "road                      1294          36       0           1382   \n",
       "cultivation               4477         126       0            748   \n",
       "slash_burn                 126         209       0             33   \n",
       "cloudy                       0           0    2089              0   \n",
       "partly_cloudy              748          33       0           7261   \n",
       "conventional_mine            4           0       0             28   \n",
       "bare_ground                 89          10       0             74   \n",
       "artisinal_mine              18           0       0             27   \n",
       "blooming                    35           2       0             17   \n",
       "selective_logging           58           2       0             27   \n",
       "blow_down                    8           2       0             13   \n",
       "\n",
       "                   conventional_mine  bare_ground  artisinal_mine  blooming  \\\n",
       "haze                               2           41               5         4   \n",
       "primary                           94          683             324       332   \n",
       "agriculture                       24          225              38        32   \n",
       "clear                             70          747             307       311   \n",
       "water                             26          206             299        16   \n",
       "habitation                        36          163              29         4   \n",
       "road                              59          323             110        10   \n",
       "cultivation                        4           89              18        35   \n",
       "slash_burn                         0           10               0         2   \n",
       "cloudy                             0            0               0         0   \n",
       "partly_cloudy                     28           74              27        17   \n",
       "conventional_mine                100           10               4         0   \n",
       "bare_ground                       10          862              40         3   \n",
       "artisinal_mine                     4           40             339         0   \n",
       "blooming                           0            3               0       332   \n",
       "selective_logging                  0           13               6         7   \n",
       "blow_down                          0            4               0         1   \n",
       "\n",
       "                   selective_logging  blow_down  \n",
       "haze                               5          0  \n",
       "primary                          340         98  \n",
       "agriculture                       65         22  \n",
       "clear                            308         85  \n",
       "water                             49          3  \n",
       "habitation                        13          3  \n",
       "road                             151          2  \n",
       "cultivation                       58          8  \n",
       "slash_burn                         2          2  \n",
       "cloudy                             0          0  \n",
       "partly_cloudy                     27         13  \n",
       "conventional_mine                  0          0  \n",
       "bare_ground                       13          4  \n",
       "artisinal_mine                     6          0  \n",
       "blooming                           7          1  \n",
       "selective_logging                340          1  \n",
       "blow_down                          1         98  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "def make_cooccurence_matrix(labels):\n",
    "    numeric_df = train_data[labels]; \n",
    "    c_matrix = numeric_df.T.dot(numeric_df)\n",
    "    sns.heatmap(c_matrix)\n",
    "    return c_matrix\n",
    "    \n",
    "# Compute the co-ocurrence matrix\n",
    "make_cooccurence_matrix(label_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.image import imread\n",
    "# load image pixels\n",
    "image = imread(train_img_path+\"/\"+\"train_10002.jpg\")\n",
    "\t# plot raw pixel data\n",
    "plt.imshow(image)\n",
    "    \n",
    "# show the figure\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(256, 256, 4)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "from keras.preprocessing.image import load_img\n",
    "from keras.preprocessing.image import img_to_array\n",
    "from os import listdir\n",
    "from numpy import zeros\n",
    "from numpy import asarray\n",
    "from numpy import savez_compressed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def create_tag_mapping(labels_df):\n",
    "    \n",
    "    labels = set()\n",
    "    for i in range(len(labels_df)):\n",
    "        # convert spaced separated tags into an array of tags\n",
    "        tags = labels_df['tags'][i].split(' ')\n",
    "        # add tags to the set of known labels\n",
    "        labels.update(tags)\n",
    "    # convert set of labels to a list to list\n",
    "    labels = list(labels)\n",
    "    labels.sort()\n",
    " # dict that maps labels to integers, and the reverse\n",
    "    labels_map = { labels [i]:i for i in range(len(labels))}\n",
    "    inv_labels_map = {i: labels [i] for i in range(len(labels))}\n",
    "    return labels_map, inv_labels_map\n",
    "\n",
    "\n",
    "\n",
    "def create_file_mapping(train_data):\n",
    "    mapping = dict()\n",
    "    for i in range(len(labels_df)):\n",
    "        name, tags = train_data['image_name'][i], labels_df['tags'][i]\n",
    "        mapping[name] = tags.split(' ')\n",
    "        \n",
    "    return mapping\n",
    "\n",
    "\n",
    "def one_hot_encode(tags,mapping):\n",
    "    \n",
    "    # create empty vector\n",
    "    encoding = zeros(len(mapping), dtype='uint8')\n",
    "    # mark 1 for each tag in the vector\n",
    "    for tag in tags:\n",
    "        encoding[mapping[tag]] = 1\n",
    "    return encoding\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "def load_dataset(path, file_mapping, tag_mapping):\n",
    "    photos, targets = list(), list()\n",
    "    # enumerate files in the directory\n",
    "    for filename in listdir(folder):\n",
    "        # load image\n",
    "        photo = load_img(path + filename, target_size=(64,64))\n",
    "        # convert to numpy array\n",
    "        photo = img_to_array(photo, dtype='uint8')\n",
    "        # get tags\n",
    "        tags = file_mapping[filename[:-4]]\n",
    "        # one hot encode tags\n",
    "        target = one_hot_encode(tags, tag_mapping)\n",
    "        # store\n",
    "        photos.append(photo)\n",
    "        targets.append(target)\n",
    "    X = asarray(photos, dtype='uint8')\n",
    "    y = asarray(targets, dtype='uint8')\n",
    "    return X, y\n",
    " \n",
    "    \n",
    "    \n",
    " \n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "17\n",
      "{'agriculture': 0, 'artisinal_mine': 1, 'bare_ground': 2, 'blooming': 3, 'blow_down': 4, 'clear': 5, 'cloudy': 6, 'conventional_mine': 7, 'cultivation': 8, 'habitation': 9, 'haze': 10, 'partly_cloudy': 11, 'primary': 12, 'road': 13, 'selective_logging': 14, 'slash_burn': 15, 'water': 16}\n",
      "(40479, 64, 64, 3) (40479, 17)\n"
     ]
    }
   ],
   "source": [
    "filename = '../input/planets-dataset/planet/planet/train_classes.csv'\n",
    "labels_df = pd.read_csv(filename, encoding='latin1')\n",
    "\n",
    "# create a mapping of tags to integers\n",
    "mapping, inv_mapping = create_tag_mapping(labels_df)\n",
    "print(len(mapping))\n",
    "print(mapping)\n",
    "\n",
    "# create a mapping of tags to integers\n",
    "tag_mapping, _ = create_tag_mapping(labels_df)\n",
    "\n",
    "# create a mapping of filenames to tag lists\n",
    "file_mapping = create_file_mapping(labels_df)\n",
    "\n",
    "# load the jpeg images\n",
    "folder = '../input/planets-dataset/planet/planet/train-jpg/'\n",
    "X, y = load_dataset(folder, file_mapping, tag_mapping)\n",
    "print(X.shape, y.shape)\n",
    "\n",
    "# save both arrays to one file in compressed format\n",
    "savez_compressed('planet_data.npz', X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import keras \n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt \n",
    "from keras.models import Sequential \n",
    "from keras.layers import Conv2D,Flatten,MaxPooling2D,Dropout,Dense,BatchNormalization,SpatialDropout2D\n",
    "from keras.applications.vgg16 import VGG16\n",
    "from keras.models import Model\n",
    "from keras.optimizers import SGD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[[ 70,  90,  68],\n",
       "         [ 77,  91,  72],\n",
       "         [ 73,  90,  74],\n",
       "         ...,\n",
       "         [ 42,  56,  48],\n",
       "         [ 50,  63,  54],\n",
       "         [ 47,  66,  56]],\n",
       "\n",
       "        [[ 74,  88,  73],\n",
       "         [ 75,  89,  72],\n",
       "         [ 74,  89,  70],\n",
       "         ...,\n",
       "         [ 33,  51,  49],\n",
       "         [ 49,  66,  55],\n",
       "         [ 41,  60,  50]],\n",
       "\n",
       "        [[ 75,  90,  73],\n",
       "         [ 75,  90,  73],\n",
       "         [ 75,  91,  73],\n",
       "         ...,\n",
       "         [ 42,  59,  51],\n",
       "         [ 46,  61,  53],\n",
       "         [ 41,  64,  52]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[159, 139, 114],\n",
       "         [155, 136, 113],\n",
       "         [145, 130, 105],\n",
       "         ...,\n",
       "         [ 76,  93,  75],\n",
       "         [ 75,  92,  72],\n",
       "         [ 80,  92,  76]],\n",
       "\n",
       "        [[157, 139, 118],\n",
       "         [152, 134, 114],\n",
       "         [134, 123,  99],\n",
       "         ...,\n",
       "         [ 80,  91,  73],\n",
       "         [ 77,  90,  71],\n",
       "         [ 84,  90,  75]],\n",
       "\n",
       "        [[153, 139, 115],\n",
       "         [159, 140, 118],\n",
       "         [146, 131, 105],\n",
       "         ...,\n",
       "         [ 83,  95,  79],\n",
       "         [ 75,  90,  72],\n",
       "         [ 80,  90,  75]]],\n",
       "\n",
       "\n",
       "       [[[141, 126, 109],\n",
       "         [136, 123, 106],\n",
       "         [136, 122, 105],\n",
       "         ...,\n",
       "         [140, 127, 111],\n",
       "         [141, 127, 112],\n",
       "         [138, 125, 109]],\n",
       "\n",
       "        [[139, 125, 108],\n",
       "         [139, 126, 109],\n",
       "         [136, 123, 105],\n",
       "         ...,\n",
       "         [139, 127, 109],\n",
       "         [143, 129, 111],\n",
       "         [143, 129, 114]],\n",
       "\n",
       "        [[138, 129, 109],\n",
       "         [138, 126, 109],\n",
       "         [139, 124, 108],\n",
       "         ...,\n",
       "         [130, 118, 104],\n",
       "         [133, 122, 104],\n",
       "         [138, 127, 109]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[101, 102,  91],\n",
       "         [114, 111,  98],\n",
       "         [115, 114, 100],\n",
       "         ...,\n",
       "         [110, 110,  96],\n",
       "         [118, 114,  99],\n",
       "         [102, 107,  92]],\n",
       "\n",
       "        [[108, 104,  95],\n",
       "         [107, 111,  96],\n",
       "         [103, 108,  92],\n",
       "         ...,\n",
       "         [117, 113,  97],\n",
       "         [ 97, 103,  89],\n",
       "         [107, 107,  96]],\n",
       "\n",
       "        [[109, 106,  95],\n",
       "         [109, 111,  95],\n",
       "         [109, 113,  99],\n",
       "         ...,\n",
       "         [105, 105,  93],\n",
       "         [109, 109,  93],\n",
       "         [107, 106,  93]]],\n",
       "\n",
       "\n",
       "       [[[ 22,  34,  24],\n",
       "         [ 28,  40,  27],\n",
       "         [ 21,  34,  24],\n",
       "         ...,\n",
       "         [ 30,  42,  26],\n",
       "         [ 28,  39,  28],\n",
       "         [ 29,  42,  28]],\n",
       "\n",
       "        [[ 29,  38,  25],\n",
       "         [ 35,  46,  30],\n",
       "         [ 31,  43,  28],\n",
       "         ...,\n",
       "         [ 31,  42,  27],\n",
       "         [ 35,  44,  28],\n",
       "         [ 23,  37,  28]],\n",
       "\n",
       "        [[ 27,  37,  26],\n",
       "         [ 38,  47,  32],\n",
       "         [ 34,  42,  29],\n",
       "         ...,\n",
       "         [ 29,  42,  29],\n",
       "         [ 25,  38,  26],\n",
       "         [ 34,  42,  28]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[ 40,  46,  31],\n",
       "         [ 32,  42,  29],\n",
       "         [ 30,  39,  27],\n",
       "         ...,\n",
       "         [ 24,  37,  26],\n",
       "         [ 35,  45,  29],\n",
       "         [ 34,  46,  30]],\n",
       "\n",
       "        [[ 23,  39,  25],\n",
       "         [ 30,  42,  29],\n",
       "         [ 32,  43,  28],\n",
       "         ...,\n",
       "         [ 23,  37,  25],\n",
       "         [ 22,  32,  27],\n",
       "         [ 33,  41,  31]],\n",
       "\n",
       "        [[ 29,  40,  28],\n",
       "         [ 23,  36,  22],\n",
       "         [ 33,  46,  29],\n",
       "         ...,\n",
       "         [ 25,  35,  25],\n",
       "         [ 31,  41,  29],\n",
       "         [ 27,  34,  24]]],\n",
       "\n",
       "\n",
       "       ...,\n",
       "\n",
       "\n",
       "       [[[125, 130, 122],\n",
       "         [126, 129, 122],\n",
       "         [127, 131, 122],\n",
       "         ...,\n",
       "         [120, 125, 116],\n",
       "         [122, 126, 114],\n",
       "         [119, 125, 117]],\n",
       "\n",
       "        [[123, 129, 119],\n",
       "         [126, 133, 121],\n",
       "         [128, 132, 123],\n",
       "         ...,\n",
       "         [119, 124, 116],\n",
       "         [116, 124, 118],\n",
       "         [117, 125, 115]],\n",
       "\n",
       "        [[126, 131, 120],\n",
       "         [124, 130, 120],\n",
       "         [127, 131, 124],\n",
       "         ...,\n",
       "         [120, 126, 117],\n",
       "         [117, 126, 115],\n",
       "         [120, 123, 116]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[125, 130, 122],\n",
       "         [127, 132, 123],\n",
       "         [127, 133, 125],\n",
       "         ...,\n",
       "         [120, 127, 117],\n",
       "         [122, 125, 118],\n",
       "         [120, 128, 119]],\n",
       "\n",
       "        [[126, 131, 124],\n",
       "         [123, 130, 123],\n",
       "         [128, 133, 126],\n",
       "         ...,\n",
       "         [122, 127, 118],\n",
       "         [123, 127, 120],\n",
       "         [121, 125, 119]],\n",
       "\n",
       "        [[127, 132, 123],\n",
       "         [128, 132, 120],\n",
       "         [125, 130, 121],\n",
       "         ...,\n",
       "         [118, 125, 117],\n",
       "         [123, 128, 117],\n",
       "         [121, 128, 116]]],\n",
       "\n",
       "\n",
       "       [[[197, 197, 184],\n",
       "         [197, 197, 184],\n",
       "         [197, 197, 184],\n",
       "         ...,\n",
       "         [196, 195, 183],\n",
       "         [194, 194, 183],\n",
       "         [194, 194, 183]],\n",
       "\n",
       "        [[194, 195, 181],\n",
       "         [194, 195, 181],\n",
       "         [194, 194, 181],\n",
       "         ...,\n",
       "         [194, 195, 183],\n",
       "         [194, 194, 181],\n",
       "         [194, 194, 181]],\n",
       "\n",
       "        [[193, 193, 180],\n",
       "         [193, 193, 180],\n",
       "         [193, 194, 180],\n",
       "         ...,\n",
       "         [193, 193, 182],\n",
       "         [192, 193, 180],\n",
       "         [192, 193, 180]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[180, 182, 170],\n",
       "         [180, 180, 170],\n",
       "         [180, 181, 169],\n",
       "         ...,\n",
       "         [176, 178, 166],\n",
       "         [177, 178, 166],\n",
       "         [177, 178, 166]],\n",
       "\n",
       "        [[180, 182, 169],\n",
       "         [180, 182, 169],\n",
       "         [180, 181, 170],\n",
       "         ...,\n",
       "         [177, 179, 166],\n",
       "         [177, 179, 165],\n",
       "         [177, 179, 165]],\n",
       "\n",
       "        [[180, 180, 167],\n",
       "         [180, 180, 167],\n",
       "         [180, 180, 168],\n",
       "         ...,\n",
       "         [177, 179, 166],\n",
       "         [177, 179, 165],\n",
       "         [177, 179, 165]]],\n",
       "\n",
       "\n",
       "       [[[ 56,  69,  63],\n",
       "         [ 56,  69,  64],\n",
       "         [ 61,  72,  67],\n",
       "         ...,\n",
       "         [ 48,  63,  61],\n",
       "         [ 61,  73,  66],\n",
       "         [ 57,  72,  66]],\n",
       "\n",
       "        [[ 55,  70,  64],\n",
       "         [ 57,  71,  67],\n",
       "         [ 56,  75,  66],\n",
       "         ...,\n",
       "         [ 49,  61,  57],\n",
       "         [ 67,  79,  68],\n",
       "         [ 59,  71,  67]],\n",
       "\n",
       "        [[ 55,  65,  53],\n",
       "         [ 53,  66,  64],\n",
       "         [ 56,  67,  70],\n",
       "         ...,\n",
       "         [ 55,  72,  64],\n",
       "         [ 63,  73,  69],\n",
       "         [ 56,  71,  61]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[ 51,  66,  59],\n",
       "         [ 63,  73,  67],\n",
       "         [ 54,  67,  62],\n",
       "         ...,\n",
       "         [ 55,  73,  61],\n",
       "         [ 46,  61,  59],\n",
       "         [ 52,  66,  60]],\n",
       "\n",
       "        [[ 53,  65,  60],\n",
       "         [ 57,  70,  65],\n",
       "         [ 53,  64,  62],\n",
       "         ...,\n",
       "         [ 55,  72,  66],\n",
       "         [ 50,  67,  63],\n",
       "         [ 52,  66,  63]],\n",
       "\n",
       "        [[ 50,  67,  62],\n",
       "         [ 52,  64,  59],\n",
       "         [ 51,  69,  62],\n",
       "         ...,\n",
       "         [ 51,  67,  63],\n",
       "         [ 54,  67,  62],\n",
       "         [ 51,  66,  59]]]], dtype=uint8)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = np.load('./planet_data.npz')\n",
    "X,y =data['arr_0'],data['arr_1']\n",
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import backend\n",
    "\n",
    "\n",
    "def fbeta(y_true, y_pred, beta=2):\n",
    "    # clip predictions\n",
    "    y_pred = backend.clip(y_pred, 0, 1)\n",
    "    # calculate elements\n",
    "    tp = backend.sum(backend.round(backend.clip(y_true * y_pred, 0, 1)), axis=1)\n",
    "    fp = backend.sum(backend.round(backend.clip(y_pred - y_true, 0, 1)), axis=1)\n",
    "    fn = backend.sum(backend.round(backend.clip(y_true - y_pred, 0, 1)), axis=1)\n",
    "    # calculate precision\n",
    "    p = tp / (tp + fp + backend.epsilon())\n",
    "    # calculate recall\n",
    "    r = tp / (tp + fn + backend.epsilon())\n",
    "    # calculate fbeta, averaged across each class\n",
    "    bb = beta ** 2\n",
    "    fbeta_score = backend.mean((1 + bb) * (p * r) / (bb * p + r + backend.epsilon()))\n",
    "    return fbeta_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "\n",
    "train_datagen = ImageDataGenerator(featurewise_center=True, horizontal_flip=True, vertical_flip=True, rotation_range=90)\n",
    "test_datagen = ImageDataGenerator(featurewise_center=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28335, 64, 64, 3) (28335, 17) (12144, 64, 64, 3) (12144, 17)\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, Y_train, Y_test = train_test_split(X, y, test_size=0.3, random_state=1)\n",
    "\n",
    "print(X_train.shape, Y_train.shape, X_test.shape, Y_test.shape)\n",
    "\n",
    "train_datagen.mean = [123.68, 116.779, 103.939]\n",
    "test_datagen.mean = [123.68, 116.779, 103.939]\n",
    "train_datagen.fit(X_train)\n",
    "test_datagen.fit(X_test)\n",
    "\n",
    "train_it = train_datagen.flow(X_train, Y_train, batch_size=128)\n",
    "\n",
    "test_it = test_datagen.flow(X_test, Y_test, batch_size=128)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model():\n",
    "    in_shape=(64, 64, 3)\n",
    "    out_shape=17\n",
    "\n",
    "    # load mannequin\n",
    "\n",
    "    mannequin = VGG16(include_top=False, input_shape=in_shape)\n",
    "\n",
    "    # mark loaded layers as not trainable\n",
    "\n",
    "    for layer in mannequin.layers:\n",
    "        layer.trainable = False\n",
    "\n",
    "    # enable final vgg block to be trainable\n",
    "\n",
    "    mannequin.get_layer(\"block5_conv1\").trainable = True\n",
    "\n",
    "    mannequin.get_layer(\"block5_conv2\").trainable = True\n",
    "\n",
    "    mannequin.get_layer(\"block5_conv3\").trainable = True\n",
    "\n",
    "    mannequin.get_layer(\"block5_pool\").trainable = True\n",
    "\n",
    "    # add new classifier layers\n",
    "\n",
    "    flat1 = Flatten()(mannequin.layers[-1].output)\n",
    "\n",
    "    class1 = Dense(128, activation=\"relu\", kernel_initializer=\"he_uniform\")(flat1)\n",
    "\n",
    "    output = Dense(out_shape, activation=\"sigmoid\")(class1)\n",
    "\n",
    "    # outline new mannequin\n",
    "\n",
    "    mannequin = Model(inputs=mannequin.inputs, outputs=output)\n",
    "\n",
    "    # compile mannequin\n",
    "\n",
    "    choose = SGD(lr=0.01, momentum=0.9)\n",
    "\n",
    "    mannequin.compile(optimizer=choose, loss=\"binary_crossentropy\", metrics=[fbeta])\n",
    "    \n",
    "    return mannequin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "222/222 [==============================] - 43s 192ms/step - loss: 0.1183 - fbeta: 0.8621 - val_loss: 0.1221 - val_fbeta: 0.8559\n",
      "Epoch 2/10\n",
      "222/222 [==============================] - 42s 190ms/step - loss: 0.1158 - fbeta: 0.8654 - val_loss: 0.1285 - val_fbeta: 0.8524\n",
      "Epoch 3/10\n",
      "222/222 [==============================] - 42s 188ms/step - loss: 0.1146 - fbeta: 0.8670 - val_loss: 0.1275 - val_fbeta: 0.8489\n",
      "Epoch 4/10\n",
      "222/222 [==============================] - 43s 192ms/step - loss: 0.1126 - fbeta: 0.8696 - val_loss: 0.1243 - val_fbeta: 0.8500\n",
      "Epoch 5/10\n",
      "222/222 [==============================] - 43s 195ms/step - loss: 0.1118 - fbeta: 0.8688 - val_loss: 0.1184 - val_fbeta: 0.8600\n",
      "Epoch 6/10\n",
      "222/222 [==============================] - 42s 188ms/step - loss: 0.1104 - fbeta: 0.8709 - val_loss: 0.1196 - val_fbeta: 0.8535\n",
      "Epoch 7/10\n",
      "222/222 [==============================] - 43s 192ms/step - loss: 0.1097 - fbeta: 0.8732 - val_loss: 0.1196 - val_fbeta: 0.8617\n",
      "Epoch 8/10\n",
      "222/222 [==============================] - 43s 195ms/step - loss: 0.1083 - fbeta: 0.8740 - val_loss: 0.1233 - val_fbeta: 0.8582\n",
      "Epoch 9/10\n",
      "222/222 [==============================] - 43s 192ms/step - loss: 0.1076 - fbeta: 0.8743 - val_loss: 0.1224 - val_fbeta: 0.8543\n",
      "Epoch 10/10\n",
      "222/222 [==============================] - 41s 186ms/step - loss: 0.1078 - fbeta: 0.8748 - val_loss: 0.1200 - val_fbeta: 0.8536\n",
      "> loss=0.120, fbeta=0.854\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.python.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "\n",
    "# cb_early_stopper = EarlyStopping(monitor = 'val_loss', patience = EARLY_STOP_PATIENCE)\n",
    "# cb_checkpointer = ModelCheckpoint(filepath = '../working/best.hdf5', monitor = 'val_loss', save_best_only = True, mode = 'auto')\n",
    "save_best_check_point = ModelCheckpoint(filepath='our_model.hdf5', monitor='val_fbeta', mode='max', save_best_only=True, save_weights_only=True)\n",
    "# checkpoint = ModelCheckpoint(filepath='weights/weights.hdf5',monitor='val_fbeta', verbose=1, save_best_only=True, save_weights_only=False, mode='auto', period=1)\n",
    "# early = EarlyStopping(monitor='val_fbeta', min_delta=0, patience=20, verbose=1, mode='auto')\n",
    "\n",
    "# fit model\n",
    "# history = full_model.fit_generator(train_it, steps_per_epoch=len(train_it),validation_data=test_it, validation_steps=len(test_it), epochs=20, verbose=0)\n",
    "mannequin.fit_generator(train_it, steps_per_epoch=len(train_it),validation_data=test_it, validation_steps=len(test_it), epochs=10,callbacks=[save_best_check_point])\n",
    "\n",
    "\n",
    "# evaluate model\n",
    "loss, fbeta = mannequin.evaluate_generator(test_it, steps=len(test_it), verbose=0)\n",
    "print('> loss=%.3f, fbeta=%.3f' % (loss, fbeta))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plt.figure(1, figsize = (8,8)) \n",
    "# plt.subplot(211)\n",
    "# plt.title('Cross Entropy Loss')\n",
    "# plt.plot(history.history['loss'], color='blue', label='train')\n",
    "# plt.plot(history.history['val_loss'], color='orange', label='test')\n",
    "# plt.ylabel('loss')\n",
    "# plt.xlabel('epoch')\n",
    "# plt.legend(['train', 'valid']) \n",
    "\n",
    "# # plot accuracy\n",
    "# plt.subplot(212)\n",
    "# plt.title('Fbeta')\n",
    "# plt.plot(history.history['fbeta'], color='blue', label='train')\n",
    "# plt.plot(history.history['val_fbeta'], color='orange', label='test')\n",
    "# plt.ylabel('accuracy')\n",
    "# plt.xlabel('epoch')\n",
    "# plt.legend(['train', 'valid'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__notebook_source__.ipynb  best_model.hdf5  our_model.hdf5  planet_data.npz\r\n"
     ]
    }
   ],
   "source": [
    "! ls ./"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "model2 = build_model() # building a sequential model for testing\n",
    "\n",
    "#loading in the weights of the trained model\n",
    "model2.load_weights('best_model.hdf5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test_0.jpg</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test_1.jpg</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test_2.jpg</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test_3.jpg</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test_4.jpg</td>\n",
       "      <td>primary clear agriculture road water</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   image_name                                  tags\n",
       "0  test_0.jpg  primary clear agriculture road water\n",
       "1  test_1.jpg  primary clear agriculture road water\n",
       "2  test_2.jpg  primary clear agriculture road water\n",
       "3  test_3.jpg  primary clear agriculture road water\n",
       "4  test_4.jpg  primary clear agriculture road water"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_submission = sample_submission_df.copy()\n",
    "sample_submission['image_name'] = sample_submission['image_name'].apply(lambda x: '{}.jpg'.format(x))\n",
    "sample_submission.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 40669 validated image filenames.\n"
     ]
    }
   ],
   "source": [
    "test1_df = sample_submission.iloc[:40669]['image_name'].reset_index().drop('index', axis=1)\n",
    "test1_df.head()\n",
    "test1_df.shape\n",
    "# initializing an image data generator object for the first 40669 images in the sample submission dataframe\n",
    "test_image_gen1 = tf.keras.preprocessing.image.ImageDataGenerator(rescale=1/255)\n",
    "\n",
    "# generating the image data for the first 40669 images in the sample submission dataframe\n",
    "test_gen1 = test_image_gen1.flow_from_dataframe(dataframe=test1_df, directory='../input/planets-dataset/planet/planet/test-jpg/', x_col='image_name', y_col=None, batch_size=128, shuffle=False, class_mode=None, target_size=(64,64))\n",
    "\n",
    "# setting the step size for the testing set for the first 40669 images in the sample submission dataframe\n",
    "step_test_size1 = int(np.ceil(test_gen1.samples / test_gen1.batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "318/318 [==============================] - 127s 398ms/step\n"
     ]
    }
   ],
   "source": [
    "test_gen1.reset() # reseting the generator to be sure of avoiding shuffling\n",
    "pred1 = model2.predict(test_gen1, steps=step_test_size1, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test_0.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test_1.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test_2.jpg</td>\n",
       "      <td>partly_cloudy primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test_3.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test_4.jpg</td>\n",
       "      <td>partly_cloudy primary</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   image_name                   tags\n",
       "0  test_0.jpg          clear primary\n",
       "1  test_1.jpg          clear primary\n",
       "2  test_2.jpg  partly_cloudy primary\n",
       "3  test_3.jpg          clear primary\n",
       "4  test_4.jpg  partly_cloudy primary"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_file_names1 = test_gen1.filenames # storing the filenames (images names) of the first 40669 images names in \n",
    "                                       # the sample submission dataframe as ordered in the prediction as a \n",
    "                                       # variable\n",
    "\n",
    "# converting the predictions of the first 40669 to tag names\n",
    "pred_tags1 = pd.DataFrame(pred1)\n",
    "pred_tags1 = pred_tags1.apply(lambda x: ' '.join(np.array(unique_labels)[x > 0.5]), axis=1)\n",
    "\n",
    "# converting the predictions of the first 40669 to a dataframe\n",
    "result1 = pd.DataFrame({'image_name': test_file_names1, 'tags': pred_tags1})\n",
    "result1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>file_0.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>file_1.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>file_10.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>file_100.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>file_1000.jpg</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      image_name\n",
       "0     file_0.jpg\n",
       "1     file_1.jpg\n",
       "2    file_10.jpg\n",
       "3   file_100.jpg\n",
       "4  file_1000.jpg"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test2_df = sample_submission.iloc[40669:]['image_name'].reset_index().drop('index', axis=1)\n",
    "test2_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 20522 validated image filenames.\n"
     ]
    }
   ],
   "source": [
    "test_image_gen2 = tf.keras.preprocessing.image.ImageDataGenerator(rescale=1/255)\n",
    "\n",
    "# generating the image data for the remaining images in the sample submission dataframe\n",
    "test_gen2 = test_image_gen2.flow_from_dataframe(dataframe=test2_df, \\\n",
    "            directory='../input/planets-dataset/test-jpg-additional/test-jpg-additional/', x_col='image_name', \\\n",
    "            y_col=None, batch_size=128, shuffle=False, class_mode=None, target_size=(64,64))\n",
    "\n",
    "# setting the step size for the testing set for the remaining images in the sample submission dataframe\n",
    "step_test_size2 = int(np.ceil(test_gen2.samples / test_gen2.batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "161/161 [==============================] - 52s 323ms/step\n"
     ]
    }
   ],
   "source": [
    "test_gen2.reset() # reseting the generator to be sure of avoiding shuffling\n",
    "pred2 = model2.predict(test_gen2, steps=step_test_size2, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>file_0.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>file_1.jpg</td>\n",
       "      <td>partly_cloudy agriculture primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>file_10.jpg</td>\n",
       "      <td>haze primary water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>file_100.jpg</td>\n",
       "      <td>clear agriculture primary water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>file_1000.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      image_name                               tags\n",
       "0     file_0.jpg                      clear primary\n",
       "1     file_1.jpg  partly_cloudy agriculture primary\n",
       "2    file_10.jpg                 haze primary water\n",
       "3   file_100.jpg    clear agriculture primary water\n",
       "4  file_1000.jpg                      clear primary"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_file_names2 = test_gen2.filenames # storing the filenames (images names) of the remaining images names in \n",
    "                                       # the sample submission dataframe as ordered in the prediction as a \n",
    "                                       # variable\n",
    "\n",
    "# converting the predictions of the remaining images to tag names\n",
    "pred_tags2 = pd.DataFrame(pred2)\n",
    "pred_tags2 = pred_tags2.apply(lambda x: ' '.join(np.array(unique_labels)[x > 0.5]), axis=1)\n",
    "\n",
    "# converting the predictions of the remaining to a dataframe\n",
    "\n",
    "len(test_file_names2)\n",
    "len(pred_tags2)\n",
    "result2 = pd.DataFrame({'image_name': test_file_names2, 'tags': pred_tags2})\n",
    "result2.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(61191, 2)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test_0.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test_1.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test_2.jpg</td>\n",
       "      <td>partly_cloudy primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test_3.jpg</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test_4.jpg</td>\n",
       "      <td>partly_cloudy primary</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   image_name                   tags\n",
       "0  test_0.jpg          clear primary\n",
       "1  test_1.jpg          clear primary\n",
       "2  test_2.jpg  partly_cloudy primary\n",
       "3  test_3.jpg          clear primary\n",
       "4  test_4.jpg  partly_cloudy primary"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_result = pd.concat([result1, result2]) # concatenate the predictions of the test.jpg and \n",
    "                                             # test-additional.jpg into a single dataframe\n",
    "\n",
    "final_result = final_result.reset_index().drop('index', axis=1) # reseting the index of the dataframe so it \n",
    "                                                                # matches that of sample submission datafarme\n",
    "\n",
    "print(final_result.shape)\n",
    "final_result.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert sum(sample_submission['image_name'] == final_result['image_name']) == 61191"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test_0</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test_1</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test_2</td>\n",
       "      <td>partly_cloudy primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test_3</td>\n",
       "      <td>clear primary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test_4</td>\n",
       "      <td>partly_cloudy primary</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  image_name                   tags\n",
       "0     test_0          clear primary\n",
       "1     test_1          clear primary\n",
       "2     test_2  partly_cloudy primary\n",
       "3     test_3          clear primary\n",
       "4     test_4  partly_cloudy primary"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# removing the .jpg extension from 'iamge_name' column\n",
    "final_result['image_name'] = final_result['image_name'].apply(lambda x: x[:-4])\n",
    "final_result.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_result.to_csv('final_submission.csv', index=False) # saving the predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}