01_data_understanding_&_cleaning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Understanding"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The below image data was collected using the [NSFW Data Scraper](https://github.com/alex000kim/nsfw_data_scraper) on Github. I will be using the `explicit` `neutral`, and `sexy` For the sake of this project, I have left out the `hentai` and `anime` categories as I am more so interested in the impact the model will have on self expression, the #freethenipple movement, and racial bias. \n",
    "\n",
    "I have also webscraped Google Images for the `violent` and `breastfeeding` categories using the [webscraping .py files]('Webscraping') I created. It is difficult to source `breastfeeding` and `violent` images on Google, so there are only a few hundred images within these classes. This may pose a problem when modeling begins, since incorrectly categorizes these images will have a low impacct on model accuracy.\n",
    "\n",
    "All 5 of these categories were noisy and contained many misclassified images which jeopardized the integrity of my model. After downloading the images, I manually sorted through them and reclassified as needed. It is possible that some images may still be misclassified, given that I sifted through close to 100k photos, there was plenty of room for human error."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries\n",
    "import os, shutil, time, scipy, cv2, random\n",
    "import numpy as np\n",
    "from sklearn.utils import shuffle\n",
    "from pathlib import Path\n",
    "import matplotlib.pyplot as plt\n",
    "from PIL import Image\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Count the images in each class from the NSFW Data Scraper to check for class imbalance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explicit Images: 44657\n",
      "Sexy Images: 12492\n",
      "Neutral Images: 35299\n",
      "\n",
      "Total Images In Dataset: 92448\n"
     ]
    }
   ],
   "source": [
    "# create lists containing each img in each class\n",
    "explicit_imgs = [img for img in os.listdir('Data/explicit')] # listdir = file path\n",
    "sexy_imgs = [img for img in os.listdir('Data/sexy')]\n",
    "neutral_imgs = [img for img in os.listdir('Data/neutral')]\n",
    "\n",
    "\n",
    "# count the number of photos in each category + in entire data set\n",
    "print(f'Explicit Images: {len(explicit_imgs)}')\n",
    "print(f'Sexy Images: {len(sexy_imgs)}')\n",
    "print(f'Neutral Images: {len(neutral_imgs)}')\n",
    "\n",
    "\n",
    "print(f'\\nTotal Images In Dataset: {len(explicit_imgs + sexy_imgs + neutral_imgs)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nearly half of all photos are explicit, I will need to address the class imbalance in future models. I can do this with oversampling, downsampling, or data augmentation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train/Test/Validation Split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because my data is not split already, I need to create my own train/test/validation folders. In the following cells I iterate through the images in their original folders -- \"Data/explicit\", \"Data/neutral\", \"Data/sexy\" -- and split 70% of each class into the new train folder, 20% into the new test folder, and 10% into the new validation folder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Original file paths for \n",
    "explicit_data = 'Data/explicit'\n",
    "neutral_data = 'Data/neutral'\n",
    "sexy_data = 'Data/sexy'\n",
    "new_dir = 'Data/clean'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create paths for each class within train/test/validation folders\n",
    "train_folder = os.path.join(new_dir, 'train')\n",
    "train_explicit = os.path.join(train_folder, 'explicit')\n",
    "train_sexy = os.path.join(train_folder, 'sexy')\n",
    "train_neutral = os.path.join(train_folder, 'neutral')\n",
    "\n",
    "test_folder = os.path.join(new_dir, 'test')\n",
    "test_explicit = os.path.join(test_folder, 'explicit')\n",
    "test_sexy = os.path.join(test_folder, 'sexy')\n",
    "test_neutral = os.path.join(test_folder, 'neutral')\n",
    "\n",
    "val_folder = os.path.join(new_dir, 'validation')\n",
    "val_explicit = os.path.join(val_folder, 'explicit')\n",
    "val_sexy = os.path.join(val_folder, 'sexy')\n",
    "val_neutral = os.path.join(val_folder, 'neutral')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Iterate through \n",
    "imgs_explicit = [file for file in os.listdir(explicit_data)]\n",
    "imgs_sexy = [file for file in os.listdir(sexy_data)]\n",
    "imgs_neutral = [file for file in os.listdir(neutral_data)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# validation explicit\n",
    "imgs = imgs_explicit[:4466]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(explicit_data, img)\n",
    "    destination = os.path.join(val_explicit, img)\n",
    "    shutil.move(origin, destination)\n",
    "# test explicit\n",
    "imgs = imgs_explicit[4467:13397]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(explicit_data, img)\n",
    "    destination = os.path.join(test_explicit, img)\n",
    "    shutil.move(origin, destination)\n",
    "# train explicit\n",
    "imgs = imgs_explicit[13398:]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(explicit_data, img)\n",
    "    destination = os.path.join(train_explicit, img)\n",
    "    shutil.move(origin, destination)    \n",
    "    \n",
    "# validation sexy\n",
    "imgs = imgs_sexy[:1249]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(sexy_data, img)\n",
    "    destination = os.path.join(val_sexy, img)\n",
    "    shutil.move(origin, destination)\n",
    "# test sexy\n",
    "imgs = imgs_sexy[1250:3747]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(sexy_data, img)\n",
    "    destination = os.path.join(test_sexy, img)\n",
    "    shutil.move(origin, destination)\n",
    "# train sexy\n",
    "imgs = imgs_sexy[3748:]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(sexy_data, img)\n",
    "    destination = os.path.join(train_sexy, img)\n",
    "    shutil.move(origin, destination)\n",
    "    \n",
    "# validation neutral\n",
    "imgs = imgs_neutral[:3530]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(neutral_data, img)\n",
    "    destination = os.path.join(val_neutral, img)\n",
    "    shutil.move(origin, destination)\n",
    "# test neutral\n",
    "imgs = imgs_neutral[3531:10590]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(neutral_data, img)\n",
    "    destination = os.path.join(test_neutral, img)\n",
    "    shutil.move(origin, destination)\n",
    "# train neutral\n",
    "imgs = imgs_neutral[10591:]\n",
    "for img in imgs:\n",
    "    origin = os.path.join(neutral_data, img)\n",
    "    destination = os.path.join(train_neutral, img)\n",
    "    shutil.move(origin, destination)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preprocessing\n",
    "\n",
    "This section of preprocessing is done only on the `explicit` `neutral`, and `sexy` image classes. Images were labeled by category and sized down to (80 x 80) in size to save computational time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7faeea6aa3d0>"
      ]
     },
     "execution_count": 20,
     "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": [
    "# View one of the images\n",
    "image = plt.imread('Data/clean/train/neutral/_108255453_tv055754582.jpg')\n",
    "plt.imshow(image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Specify the data directories, the file path for the datasets\n",
    "DATADIR = 'Data/clean/train'\n",
    "DATADIR2 = 'Data/clean/test'\n",
    "DATADIR3 = 'Data/clean/validation'\n",
    "\n",
    "# Identify the 3 classes/folders within each train/test/val folder\n",
    "CATEGORIES = ['explicit', 'neutral', 'sexy']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Resize all images\n",
    "img_size = 80"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5 µs, sys: 1e+03 ns, total: 6 µs\n",
      "Wall time: 10 µs\n",
      "Number of Photos in Training Data: 53951\n"
     ]
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label training data\n",
    "%time\n",
    "training_data = []\n",
    "\n",
    "def create_training_data():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR, category)\n",
    "        class_num = CATEGORIES.index(category) # create numerical values for the target\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array = cv2.resize(img_array, (img_size, img_size)) # resize the images\n",
    "                training_data.append([new_array, class_num])\n",
    "            except Exception as e: # pass over any broken images\n",
    "                pass\n",
    "create_training_data()\n",
    "print(f'Number of Photos in Training Data: {len(training_data)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Number of Photos in Test Data: 18481'"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label test data\n",
    "test_data = []\n",
    "\n",
    "def create_test_data():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR2, category)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array2 = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array2 = cv2.resize(img_array2, (img_size, img_size))\n",
    "                test_data.append([new_array2, class_num])\n",
    "            except Exception as e:\n",
    "                pass\n",
    "create_test_data()\n",
    "(f'Number of Photos in Test Data: {len(test_data)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Number of Photos in Validation Data: 9245'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label validation data\n",
    "val_data = []\n",
    "\n",
    "def create_val_data():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR3, category)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array3 = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array3 = cv2.resize(img_array3, (img_size, img_size))\n",
    "                val_data.append([new_array3, class_num])\n",
    "            except Exception as e:\n",
    "                pass\n",
    "create_val_data()\n",
    "(f'Number of Photos in Validation Data: {len(val_data)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffle the images\n",
    "random.shuffle(training_data)\n",
    "random.shuffle(test_data)\n",
    "random.shuffle(val_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X and y\n",
    "X = []\n",
    "y = []\n",
    "\n",
    "for features, label in training_data:\n",
    "    X.append(features)\n",
    "    y.append(label)\n",
    "X = np.array(X).reshape(-1, img_size, img_size, 1)\n",
    "y = np.array(y)\n",
    "\n",
    "# Scale the data\n",
    "X = X/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X_test and y_test\n",
    "X_test = []\n",
    "y_test = []\n",
    "\n",
    "for features, label in test_data:\n",
    "    X_test.append(features)\n",
    "    y_test.append(label)\n",
    "X_test = np.array(X_test).reshape(-1, img_size, img_size, 1)\n",
    "y_test = np.array(y_test)\n",
    "\n",
    "# Scale the data\n",
    "X_test = X_test/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X_val and y_val\n",
    "X_val = []\n",
    "y_val = []\n",
    "\n",
    "for features, label in val_data:\n",
    "    X_val.append(features)\n",
    "    y_val.append(label)\n",
    "X_val = np.array(X_val).reshape(-1, img_size, img_size, 1)\n",
    "y_val = np.array(y_val)\n",
    "\n",
    "# Scale the data\n",
    "X_val = X_val/255."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that my images are preprocessed I have decided to pickle the data, so I can remove the original images from storage."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle training data\n",
    "pickle_out = open('X.pickle', 'wb')\n",
    "pickle.dump(X, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y.pickle', 'wb')\n",
    "pickle.dump(y, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle test data\n",
    "pickle_out = open('X_test.pickle', 'wb')\n",
    "pickle.dump(X_test, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_test.pickle', 'wb')\n",
    "pickle.dump(y_test, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle validation data\n",
    "pickle_out = open('X_val.pickle', 'wb')\n",
    "pickle.dump(X_val, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_val.pickle', 'wb')\n",
    "pickle.dump(y_val, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Preprocess and Clean `violent` and `breastfeeding` Image Classes "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Specify data directories, the file path for my datasets\n",
    "DATADIR4 = 'Data/webscraped_images/new_train'\n",
    "DATADIR5 = 'Data/webscraped_images/new_val'\n",
    "\n",
    "# Identify the two additional classes\n",
    "CATEGORIES = ['breastfeeding', 'violence']\n",
    "img_size = 80"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of Photos in Training Data: 569\n"
     ]
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label training data\n",
    "training_data_extra = []\n",
    "\n",
    "def create_training_data_extra():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR4, category)\n",
    "        class_num = CATEGORIES.index(category) # create numerical values for the target\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array = cv2.resize(img_array, (img_size, img_size)) # resize the images\n",
    "                training_data_extra.append([new_array, class_num])\n",
    "            except Exception as e: # pass over any broken images\n",
    "                pass\n",
    "create_training_data_extra()\n",
    "print(f'Number of Photos in Training Data: {len(training_data_extra)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Number of Photos in Validation Data: 50'"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label validation data\n",
    "val_data_extra = []\n",
    "\n",
    "def create_val_data_extra():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR5, category)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array3 = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array3 = cv2.resize(img_array3, (img_size, img_size))\n",
    "                val_data_extra.append([new_array3, class_num])\n",
    "            except Exception as e:\n",
    "                pass\n",
    "create_val_data_extra()\n",
    "(f'Number of Photos in Validation Data: {len(val_data_extra)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are only between 200-300 photos in the `violent` and `breastfeeding` categories compared the the thousands in the `explicit`, `neutral`, and `sexy` image classes. There are no test images for `violent` and `breastfeeding` because I want to spare as few images as possible in the train category."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffle train and validation data\n",
    "random.shuffle(training_data_extra)\n",
    "random.shuffle(val_data_extra)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X_extra and y_extra\n",
    "X_extra = []\n",
    "y_extra = []\n",
    "\n",
    "for features, label in training_data_extra:\n",
    "    X_extra.append(features)\n",
    "    y_extra.append(label)\n",
    "X_extra = np.array(X_extra).reshape(-1, img_size, img_size, 1)\n",
    "y_extra = np.array(y_extra)\n",
    "\n",
    "# Scale the data\n",
    "X_extra = X_extra/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X_val_extra and y_val_extra\n",
    "X_val_extra = []\n",
    "y_val_extra = []\n",
    "\n",
    "for features, label in val_data_extra:\n",
    "    X_val_extra.append(features)\n",
    "    y_val_extra.append(label)\n",
    "X_val_extra = np.array(X_val_extra).reshape(-1, img_size, img_size, 1)\n",
    "y_val_extra = np.array(y_val_extra)\n",
    "\n",
    "# Scale the data\n",
    "X_val_extra = X_val_extra/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle X_extra and y_extra\n",
    "pickle_out = open('X_extra.pickle', 'wb')\n",
    "pickle.dump(X_extra, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_extra.pickle', 'wb')\n",
    "pickle.dump(y_extra, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle X_val_extra and y_val_extra\n",
    "pickle_out = open('X_val_extra.pickle', 'wb')\n",
    "pickle.dump(X_val_extra, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_val_extra.pickle', 'wb')\n",
    "pickle.dump(y_val_extra, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Downsampled 5-Class Image Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Specify data directories, the file path for my datasets\n",
    "DATADIR6 = 'Data/clean/train_down'\n",
    "DATADIR7 = 'Data/clean/validation_down'\n",
    "\n",
    "# Identify five classes\n",
    "CATEGORIES = ['breastfeeding', 'explicit', 'neutral', 'sexy', 'violence']\n",
    "img_size = 80"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of Photos in Training Data: 1451\n"
     ]
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label training data\n",
    "training_data_extra = []\n",
    "\n",
    "def create_training_data_extra():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR6, category)\n",
    "        class_num = CATEGORIES.index(category) # create numerical values for the target\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array = cv2.resize(img_array, (img_size, img_size)) # resize the images\n",
    "                training_data_extra.append([new_array, class_num])\n",
    "            except Exception as e: # pass over any broken images\n",
    "                pass\n",
    "create_training_data_extra()\n",
    "print(f'Number of Photos in Training Data: {len(training_data_extra)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Number of Photos in Validation Data: 110'"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label validation data\n",
    "val_data_extra = []\n",
    "\n",
    "def create_val_data_extra():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR7, category)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array3 = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array3 = cv2.resize(img_array3, (img_size, img_size))\n",
    "                val_data_extra.append([new_array3, class_num])\n",
    "            except Exception as e:\n",
    "                pass\n",
    "create_val_data_extra()\n",
    "(f'Number of Photos in Validation Data: {len(val_data_extra)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffle data\n",
    "random.shuffle(training_data_extra)\n",
    "random.shuffle(val_data_extra)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X and y\n",
    "X_extra = []\n",
    "y_extra = []\n",
    "\n",
    "for features, label in training_data_extra:\n",
    "    X_extra.append(features)\n",
    "    y_extra.append(label)\n",
    "X_extra = np.array(X_extra).reshape(-1, img_size, img_size, 1)\n",
    "y_extra = np.array(y_extra)\n",
    "\n",
    "# Scale the data\n",
    "X_extra = X_extra/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X and y\n",
    "X_val_extra = []\n",
    "y_val_extra = []\n",
    "\n",
    "for features, label in val_data_extra:\n",
    "    X_val_extra.append(features)\n",
    "    y_val_extra.append(label)\n",
    "X_val_extra = np.array(X_val_extra).reshape(-1, img_size, img_size, 1)\n",
    "y_val_extra = np.array(y_val_extra)\n",
    "\n",
    "# Scale the data\n",
    "X_val_extra = X_val_extra/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle X_down and y_down\n",
    "pickle_out = open('X_down.pickle', 'wb')\n",
    "pickle.dump(X_extra, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_down.pickle', 'wb')\n",
    "pickle.dump(y_extra, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle X_val_down and y_val_down\n",
    "pickle_out = open('X_val_down.pickle', 'wb')\n",
    "pickle.dump(X_val_extra, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_val_down.pickle', 'wb')\n",
    "pickle.dump(y_val_extra, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Increase Image Size in Downsized 5-Class Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Larger image size\n",
    "img_size = 521"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of Photos in Training Data: 1451\n"
     ]
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label validation data\n",
    "training_data_extra = []\n",
    "\n",
    "def create_training_data_extra():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR6, category)\n",
    "        class_num = CATEGORIES.index(category) # create numerical values for the target\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array = cv2.resize(img_array, (img_size, img_size)) # resize the images\n",
    "                training_data_extra.append([new_array, class_num])\n",
    "            except Exception as e: # pass over any broken images\n",
    "                pass\n",
    "create_training_data_extra()\n",
    "print(f'Number of Photos in Training Data: {len(training_data_extra)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Number of Photos in Validation Data: 110'"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Resize, convert to grayscale, and label validation data\n",
    "\n",
    "val_data_extra = []\n",
    "\n",
    "def create_val_data_extra():\n",
    "    for category in CATEGORIES:\n",
    "        path = os.path.join(DATADIR7, category)\n",
    "        class_num = CATEGORIES.index(category)\n",
    "        for img in os.listdir(path):\n",
    "            try:\n",
    "                img_array3 = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE)\n",
    "                new_array3 = cv2.resize(img_array3, (img_size, img_size))\n",
    "                val_data_extra.append([new_array3, class_num])\n",
    "            except Exception as e:\n",
    "                pass\n",
    "create_val_data_extra()\n",
    "(f'Number of Photos in Validation Data: {len(val_data_extra)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffle data\n",
    "random.shuffle(training_data_extra)\n",
    "random.shuffle(val_data_extra)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X and y\n",
    "X_extra = []\n",
    "y_extra = []\n",
    "\n",
    "for features, label in training_data_extra:\n",
    "    X_extra.append(features)\n",
    "    y_extra.append(label)\n",
    "X_extra = np.array(X_extra).reshape(-1, img_size, img_size, 1)\n",
    "y_extra = np.array(y_extra)\n",
    "\n",
    "# Scale the data\n",
    "X_extra = X_extra/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seperate the features and the target variables into X and y\n",
    "X_val_extra = []\n",
    "y_val_extra = []\n",
    "\n",
    "for features, label in val_data_extra:\n",
    "    X_val_extra.append(features)\n",
    "    y_val_extra.append(label)\n",
    "X_val_extra = np.array(X_val_extra).reshape(-1, img_size, img_size, 1)\n",
    "y_val_extra = np.array(y_val_extra)\n",
    "\n",
    "# Scale the data\n",
    "X_val_extra = X_val_extra/255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle X_down_521 and y_down_521\n",
    "pickle_out = open('X_down_521.pickle', 'wb')\n",
    "pickle.dump(X_extra, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_down_521.pickle', 'wb')\n",
    "pickle.dump(y_extra, pickle_out)\n",
    "pickle_out.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pickle X_val_down_521 and y_val_down_521\n",
    "pickle_out = open('X_val_down_521.pickle', 'wb')\n",
    "pickle.dump(X_val_extra, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open('y_val_down_521.pickle', 'wb')\n",
    "pickle.dump(y_val_extra, pickle_out)\n",
    "pickle_out.close()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (learn-env)",
   "language": "python",
   "name": "learn-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}