2_creation_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "SVV0l8vEck9w"
   },
   "source": [
    "**PART 2 - CREATING DATASET**\n",
    "\n",
    "Francesca De Cola, Valentina Moretto, Valentina Zangirolami"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tJTbWvfuck98"
   },
   "source": [
    "In this notebook, we load all images for each source data and create a dataset that include them. Finally, we save a dataset into a zip file. This operation allows us to load for the following notebook only the zip file without repeating the upload of all images for each notebook."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "weR5pJWLck-C"
   },
   "source": [
    "**Import packages**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Y_e2C4p_ck-G"
   },
   "outputs": [],
   "source": [
    "import glob\n",
    "from PIL import Image\n",
    "import numpy as np\n",
    "import cv2\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "import os\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_254I7aPck-L"
   },
   "source": [
    "In the following sections, we load images and label for each dataset consider (Visgram, Faces, Our acquisition, Jaffe)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5IGMhOMlck-O"
   },
   "source": [
    "*DATASET VISGRAM*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "3eYtM5Lock-R"
   },
   "outputs": [],
   "source": [
    "#Label extraction\n",
    "import os\n",
    "labels = []\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/angry'):\n",
    "    labels.append('angry')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/disgust'):\n",
    "    labels.append('disgust')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/fear'):\n",
    "    labels.append('fear')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/happy'):\n",
    "    labels.append('happy')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/neutral'):\n",
    "    labels.append('neutral')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/sad'):\n",
    "    labels.append('sad')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/visgram/surprise'):\n",
    "    labels.append('surprise')\n",
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "T8-QcwGDck-U",
    "outputId": "132d783f-73fa-46e7-8a8c-3c0ecf38713b"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 36/36 [00:00<00:00, 43.03it/s]\n",
      "100%|██████████████████████████████████████████████████████████████████████████████████| 34/34 [00:00<00:00, 47.64it/s]\n",
      "100%|██████████████████████████████████████████████████████████████████████████████████| 33/33 [00:00<00:00, 59.79it/s]\n",
      "100%|██████████████████████████████████████████████████████████████████████████████████| 32/32 [00:00<00:00, 73.01it/s]\n",
      "100%|██████████████████████████████████████████████████████████████████████████████████| 30/30 [00:00<00:00, 53.18it/s]\n",
      "100%|██████████████████████████████████████████████████████████████████████████████████| 34/34 [00:00<00:00, 45.93it/s]\n",
      "100%|██████████████████████████████████████████████████████████████████████████████████| 34/34 [00:00<00:00, 59.79it/s]\n"
     ]
    }
   ],
   "source": [
    "#Images extraction for each label\n",
    "loc1 = 'C:/Users/valen/OneDrive/Pictures/visgram/angry'\n",
    "loc2 = 'C:/Users/valen/OneDrive/Pictures/visgram/disgust'\n",
    "loc3 = 'C:/Users/valen/OneDrive/Pictures/visgram/fear'\n",
    "loc4 = 'C:/Users/valen/OneDrive/Pictures/visgram/happy'\n",
    "loc5 = 'C:/Users/valen/OneDrive/Pictures/visgram/neutral'\n",
    "loc6 = 'C:/Users/valen/OneDrive/Pictures/visgram/sad'\n",
    "loc7 = 'C:/Users/valen/OneDrive/Pictures/visgram/surprise'\n",
    "features = []\n",
    "from tqdm import tqdm\n",
    "for i in tqdm(os.listdir(loc1)):\n",
    "    features.append(cv2.imread(os.path.join(loc1,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc2)):\n",
    "    features.append(cv2.imread(os.path.join(loc2,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc3)):\n",
    "    features.append(cv2.imread(os.path.join(loc3,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc4)):\n",
    "    features.append(cv2.imread(os.path.join(loc4,i),0))\n",
    "for i in tqdm(os.listdir(loc5)):\n",
    "    features.append(cv2.imread(os.path.join(loc5,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc6)):\n",
    "    features.append(cv2.imread(os.path.join(loc6,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc7)):\n",
    "    features.append(cv2.imread(os.path.join(loc7,i),0))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_hOC5Pbsck-d"
   },
   "source": [
    "*OUR IMAGES (ACQUISITION VIA WEBCAM)*\n",
    "\n",
    "This operation is described in: Acquisition_images.ipynb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "gdISpH-bck-i"
   },
   "outputs": [],
   "source": [
    "#Label extraction\n",
    "import os\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/angry'):\n",
    "    labels.append('angry')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/disgust'):\n",
    "    labels.append('disgust')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/fear'):\n",
    "    labels.append('fear')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/happy'):\n",
    "    labels.append('happy')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/neutral'):\n",
    "    labels.append('neutral')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/sad'):\n",
    "    labels.append('sad')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/nostre/surprise'):\n",
    "    labels.append('surprise')\n",
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "4ZFm8eo7ck-l",
    "outputId": "a31b9d6e-25a2-44a3-d0c2-f7087f237f4d"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 248/248 [00:05<00:00, 48.43it/s]\n",
      "100%|████████████████████████████████████████████████████████████████████████████████| 235/235 [00:03<00:00, 73.03it/s]\n",
      "100%|████████████████████████████████████████████████████████████████████████████████| 243/243 [00:03<00:00, 73.70it/s]\n",
      "100%|████████████████████████████████████████████████████████████████████████████████| 252/252 [00:03<00:00, 66.18it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████| 281/281 [00:00<00:00, 378.77it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████| 292/292 [00:00<00:00, 573.28it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████| 242/242 [00:00<00:00, 645.68it/s]\n"
     ]
    }
   ],
   "source": [
    "#Images extraction for each label\n",
    "loc1 = 'C:/Users/valen/OneDrive/Pictures/nostre/angry'\n",
    "loc2 = 'C:/Users/valen/OneDrive/Pictures/nostre/disgust'\n",
    "loc3 = 'C:/Users/valen/OneDrive/Pictures/nostre/fear'\n",
    "loc4 = 'C:/Users/valen/OneDrive/Pictures/nostre/happy'\n",
    "loc5 = 'C:/Users/valen/OneDrive/Pictures/nostre/neutral'\n",
    "loc6 = 'C:/Users/valen/OneDrive/Pictures/nostre/sad'\n",
    "loc7 = 'C:/Users/valen/OneDrive/Pictures/nostre/surprise'\n",
    "from tqdm import tqdm\n",
    "for i in tqdm(os.listdir(loc1)):\n",
    "    features.append(cv2.imread(os.path.join(loc1,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc2)):\n",
    "    features.append(cv2.imread(os.path.join(loc2,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc3)):\n",
    "    features.append(cv2.imread(os.path.join(loc3,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc4)):\n",
    "    features.append(cv2.imread(os.path.join(loc4,i),0))\n",
    "for i in tqdm(os.listdir(loc5)):\n",
    "    features.append(cv2.imread(os.path.join(loc5,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc6)):\n",
    "    features.append(cv2.imread(os.path.join(loc6,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc7)):\n",
    "    features.append(cv2.imread(os.path.join(loc7,i),0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5RQEIZoEck-p"
   },
   "source": [
    "*DATASET FACES*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "7firGimDck-r"
   },
   "outputs": [],
   "source": [
    "#Label extraction\n",
    "import os\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/faces_reshape/angry'):\n",
    "    labels.append('angry')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/faces_reshape/disgust'):\n",
    "    labels.append('disgust')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/faces_reshape/fear'):\n",
    "    labels.append('fear')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/faces_reshape/happy'):\n",
    "    labels.append('happy')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/faces_reshape/neutral'):\n",
    "    labels.append('neutral')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/faces_reshape/sad'):\n",
    "    labels.append('sad')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "AT0U0oBsck-t",
    "outputId": "db533103-5e59-434f-9031-ba100ffcac9a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████████████| 4/4 [00:00<00:00, 169.41it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████████| 5/5 [00:00<00:00, 226.02it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████████| 9/9 [00:00<00:00, 253.68it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 198.21it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████████| 4/4 [00:00<00:00, 201.65it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████████| 6/6 [00:00<00:00, 259.12it/s]\n"
     ]
    }
   ],
   "source": [
    "#Images extraction for each label\n",
    "loc1 = 'C:/Users/valen/OneDrive/Pictures/faces_reshape/angry'\n",
    "loc2 = 'C:/Users/valen/OneDrive/Pictures/faces_reshape/disgust'\n",
    "loc3 = 'C:/Users/valen/OneDrive/Pictures/faces_reshape/fear'\n",
    "loc4 = 'C:/Users/valen/OneDrive/Pictures/faces_reshape/happy'\n",
    "loc5 = 'C:/Users/valen/OneDrive/Pictures/faces_reshape/neutral'\n",
    "loc6 = 'C:/Users/valen/OneDrive/Pictures/faces_reshape/sad'\n",
    "features2= []\n",
    "from tqdm import tqdm\n",
    "for i in tqdm(os.listdir(loc1)):\n",
    "    features.append(cv2.imread(os.path.join(loc1,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc2)):\n",
    "    features.append(cv2.imread(os.path.join(loc2,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc3)):\n",
    "    features.append(cv2.imread(os.path.join(loc3,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc4)):\n",
    "    features.append(cv2.imread(os.path.join(loc4,i),0))\n",
    "for i in tqdm(os.listdir(loc5)):\n",
    "    features.append(cv2.imread(os.path.join(loc5,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc6)):\n",
    "    features.append(cv2.imread(os.path.join(loc6,i),0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Bs_JxYBRck-w"
   },
   "source": [
    "*DATASET JAFFE*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "tbDmi-zIck-y"
   },
   "outputs": [],
   "source": [
    "#Label extraction\n",
    "import os\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/angry'):\n",
    "    labels.append('angry')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/disgust'):\n",
    "    labels.append('disgust')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/fear'):\n",
    "    labels.append('fear')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/happy'):\n",
    "    labels.append('happy')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/neutral'):\n",
    "    labels.append('neutral')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/sad'):\n",
    "    labels.append('sad')\n",
    "for i in os.listdir('C:/Users/valen/OneDrive/Pictures/jaffe/surprise'):\n",
    "    labels.append('surprise')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "WM2vJT73ck-0",
    "outputId": "12a57de8-d27c-41cf-802c-62577f81c064"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████████| 30/30 [00:00<00:00, 366.80it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 28/28 [00:00<00:00, 632.35it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 32/32 [00:00<00:00, 785.19it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 31/31 [00:00<00:00, 663.47it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 30/30 [00:00<00:00, 630.93it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 31/31 [00:00<00:00, 642.42it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 30/30 [00:00<00:00, 644.28it/s]\n"
     ]
    }
   ],
   "source": [
    "#Images extraction for each label\n",
    "loc1 = 'C:/Users/valen/OneDrive/Pictures/jaffe/angry'\n",
    "loc2 = 'C:/Users/valen/OneDrive/Pictures/jaffe/disgust'\n",
    "loc3 = 'C:/Users/valen/OneDrive/Pictures/jaffe/fear'\n",
    "loc4 = 'C:/Users/valen/OneDrive/Pictures/jaffe/happy'\n",
    "loc5 = 'C:/Users/valen/OneDrive/Pictures/jaffe/neutral'\n",
    "loc6 = 'C:/Users/valen/OneDrive/Pictures/jaffe/sad'\n",
    "loc7 = 'C:/Users/valen/OneDrive/Pictures/jaffe/surprise'\n",
    "from tqdm import tqdm\n",
    "for i in tqdm(os.listdir(loc1)):\n",
    "    features.append(cv2.imread(os.path.join(loc1,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc2)):\n",
    "    features.append(cv2.imread(os.path.join(loc2,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc3)):\n",
    "    features.append(cv2.imread(os.path.join(loc3,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc4)):\n",
    "    features.append(cv2.imread(os.path.join(loc4,i),0))\n",
    "for i in tqdm(os.listdir(loc5)):\n",
    "    features.append(cv2.imread(os.path.join(loc5,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc6)):\n",
    "    features.append(cv2.imread(os.path.join(loc6,i),0))\n",
    "    \n",
    "for i in tqdm(os.listdir(loc7)):\n",
    "    features.append(cv2.imread(os.path.join(loc7,i),0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3-5rJca2ck-2"
   },
   "source": [
    "We create a dataset that contains all images of all source data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "sZDDKJrcck-4"
   },
   "outputs": [],
   "source": [
    "train_data=pd.DataFrame()\n",
    "train_data['labels']=labels\n",
    "train_data['img']=features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "kuabfb7kck-5",
    "outputId": "d0d71c4f-175c-42b5-d533-e00a881d6122"
   },
   "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>labels</th>\n",
       "      <th>img</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>angry</td>\n",
       "      <td>[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>angry</td>\n",
       "      <td>[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>angry</td>\n",
       "      <td>[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>angry</td>\n",
       "      <td>[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>angry</td>\n",
       "      <td>[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2264</th>\n",
       "      <td>surprise</td>\n",
       "      <td>[[92, 88, 88, 94, 102, 90, 76, 68, 40, 18, 10,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2265</th>\n",
       "      <td>surprise</td>\n",
       "      <td>[[93, 97, 97, 85, 68, 37, 7, 18, 9, 4, 2, 3, 3...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2266</th>\n",
       "      <td>surprise</td>\n",
       "      <td>[[135, 150, 159, 151, 154, 154, 159, 163, 165,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2267</th>\n",
       "      <td>surprise</td>\n",
       "      <td>[[157, 163, 168, 164, 160, 161, 164, 163, 164,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2268</th>\n",
       "      <td>surprise</td>\n",
       "      <td>[[173, 171, 168, 168, 166, 170, 161, 163, 170,...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2269 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        labels                                                img\n",
       "0        angry  [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...\n",
       "1        angry  [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...\n",
       "2        angry  [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,...\n",
       "3        angry  [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0,...\n",
       "4        angry  [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...\n",
       "...        ...                                                ...\n",
       "2264  surprise  [[92, 88, 88, 94, 102, 90, 76, 68, 40, 18, 10,...\n",
       "2265  surprise  [[93, 97, 97, 85, 68, 37, 7, 18, 9, 4, 2, 3, 3...\n",
       "2266  surprise  [[135, 150, 159, 151, 154, 154, 159, 163, 165,...\n",
       "2267  surprise  [[157, 163, 168, 164, 160, 161, 164, 163, 164,...\n",
       "2268  surprise  [[173, 171, 168, 168, 166, 170, 161, 163, 170,...\n",
       "\n",
       "[2269 rows x 2 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "4v34U7Sack_B",
    "outputId": "1269cd8f-6831-43b9-bd97-eb2c69749080"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x285edb396d0>"
      ]
     },
     "execution_count": 14,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (8,4))\n",
    "sns.countplot(train_data['labels'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Nw9kUWTbck_E"
   },
   "source": [
    "**RESHAPE IMAGE**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VrxPDUN9ck_F"
   },
   "source": [
    "We adding the color dimension and resize shape of image from (224,224) into (224,224,3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "O4bGJOBock_G"
   },
   "outputs": [],
   "source": [
    "features = np.array(features)\n",
    "X = np.stack((features,)*3, axis=-1)\n",
    "Y = np.array(labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Sl6Xnb-Bck_H"
   },
   "source": [
    "Save dataset into npz file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "iItYHCQzck_I"
   },
   "outputs": [],
   "source": [
    "np.savez('dataset_imgs', Y=Y, X=X)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "creation_dataset_2.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}