Proyek Handwritten Ver 3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Dropout\n",
    "from keras.layers import Flatten\n",
    "from keras.layers.convolutional import Conv2D\n",
    "from keras.layers.convolutional import MaxPooling2D\n",
    "from keras.utils import np_utils\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from keras.layers import MaxPool2D\n",
    "\n",
    "from keras.preprocessing.image import ImageDataGenerator\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# load dataset\n",
    "data = pd.read_csv(\"A_Z Handwritten Data.csv\").astype('float32')\n"
   ]
  },
  {
   "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>0</th>\n",
       "      <th>0.1</th>\n",
       "      <th>0.2</th>\n",
       "      <th>0.3</th>\n",
       "      <th>0.4</th>\n",
       "      <th>0.5</th>\n",
       "      <th>0.6</th>\n",
       "      <th>0.7</th>\n",
       "      <th>0.8</th>\n",
       "      <th>0.9</th>\n",
       "      <th>...</th>\n",
       "      <th>0.562</th>\n",
       "      <th>0.563</th>\n",
       "      <th>0.564</th>\n",
       "      <th>0.565</th>\n",
       "      <th>0.566</th>\n",
       "      <th>0.567</th>\n",
       "      <th>0.568</th>\n",
       "      <th>0.569</th>\n",
       "      <th>0.570</th>\n",
       "      <th>0.571</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>372037.000000</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>...</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "      <td>372037.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>13.391642</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.762553</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>9.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>18.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>25.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0       0.1       0.2       0.3       0.4       0.5  \\\n",
       "count  372037.000000  372037.0  372037.0  372037.0  372037.0  372037.0   \n",
       "mean       13.391642       0.0       0.0       0.0       0.0       0.0   \n",
       "std         6.762553       0.0       0.0       0.0       0.0       0.0   \n",
       "min         0.000000       0.0       0.0       0.0       0.0       0.0   \n",
       "25%         9.000000       0.0       0.0       0.0       0.0       0.0   \n",
       "50%        14.000000       0.0       0.0       0.0       0.0       0.0   \n",
       "75%        18.000000       0.0       0.0       0.0       0.0       0.0   \n",
       "max        25.000000       0.0       0.0       0.0       0.0       0.0   \n",
       "\n",
       "            0.6       0.7       0.8       0.9  ...     0.562     0.563  \\\n",
       "count  372037.0  372037.0  372037.0  372037.0  ...  372037.0  372037.0   \n",
       "mean        0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "std         0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "min         0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "25%         0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "50%         0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "75%         0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "max         0.0       0.0       0.0       0.0  ...       0.0       0.0   \n",
       "\n",
       "          0.564     0.565     0.566     0.567     0.568     0.569     0.570  \\\n",
       "count  372037.0  372037.0  372037.0  372037.0  372037.0  372037.0  372037.0   \n",
       "mean        0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "std         0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "min         0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "25%         0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "50%         0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "75%         0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "max         0.0       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "\n",
       "          0.571  \n",
       "count  372037.0  \n",
       "mean        0.0  \n",
       "std         0.0  \n",
       "min         0.0  \n",
       "25%         0.0  \n",
       "50%         0.0  \n",
       "75%         0.0  \n",
       "max         0.0  \n",
       "\n",
       "[8 rows x 785 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.tail()\n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.drop('0',axis = 1) # axis=1 for dropping column (untuk jadi datanya)\n",
    "y = data['0'] # untuk jadi label prediksi\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "X_baru = X.copy()\n",
    "for i in range(len(X_baru)):\n",
    "    row = X_baru.iloc[i] # current row\n",
    "    newrow = np.array(row.values.tolist()).reshape(28,28) #jadikan 2d\n",
    "    if random.randint(0,1) == 0:\n",
    "        newrow = np.flipud(newrow).flatten() # flip vertical\n",
    "    else:\n",
    "        newrow = np.fliplr(newrow).flatten() # flip horizontal\n",
    "    X_baru.iloc[i] = newrow # override linenya lagi\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#buat column baru buat flipped A-Z\n",
    "y_baru = y.copy()\n",
    "for i in range(len(y_baru)):\n",
    "    y_baru[i] += 26\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0          0.0\n",
      "1          0.0\n",
      "2          0.0\n",
      "3          0.0\n",
      "4          0.0\n",
      "          ... \n",
      "744069    51.0\n",
      "744070    51.0\n",
      "744071    51.0\n",
      "744072    51.0\n",
      "744073    51.0\n",
      "Name: 0, Length: 744074, dtype: float32\n"
     ]
    }
   ],
   "source": [
    "#append data\n",
    "X = X.append(X_baru, ignore_index=True)\n",
    "y = y.append(y_baru, ignore_index=True)\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: (744074, 784)\n",
      "colums count: 784\n",
      "784 = 28 X 28\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>0.1</th>\n",
       "      <th>0.2</th>\n",
       "      <th>0.3</th>\n",
       "      <th>0.4</th>\n",
       "      <th>0.5</th>\n",
       "      <th>0.6</th>\n",
       "      <th>0.7</th>\n",
       "      <th>0.8</th>\n",
       "      <th>0.9</th>\n",
       "      <th>0.10</th>\n",
       "      <th>...</th>\n",
       "      <th>0.562</th>\n",
       "      <th>0.563</th>\n",
       "      <th>0.564</th>\n",
       "      <th>0.565</th>\n",
       "      <th>0.566</th>\n",
       "      <th>0.567</th>\n",
       "      <th>0.568</th>\n",
       "      <th>0.569</th>\n",
       "      <th>0.570</th>\n",
       "      <th>0.571</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 784 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   0.1  0.2  0.3  0.4  0.5  0.6  0.7  0.8  0.9  0.10  ...  0.562  0.563  \\\n",
       "0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   0.0  ...    0.0    0.0   \n",
       "1  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   0.0  ...    0.0    0.0   \n",
       "2  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   0.0  ...    0.0    0.0   \n",
       "3  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   0.0  ...    0.0    0.0   \n",
       "4  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   0.0  ...    0.0    0.0   \n",
       "\n",
       "   0.564  0.565  0.566  0.567  0.568  0.569  0.570  0.571  \n",
       "0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "1    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "2    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "3    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "4    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "\n",
       "[5 rows x 784 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"shape:\", X.shape)\n",
    "print(\"colums count:\", len(X.iloc[0]))\n",
    "print(\"784 = 28 X 28\")\n",
    "\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.0\n",
       "1    0.0\n",
       "2    0.0\n",
       "3    0.0\n",
       "4    0.0\n",
       "Name: 0, dtype: float32"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "744074"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Display Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.utils import shuffle\n",
    "\n",
    "X_shuffle = shuffle(X)\n",
    "\n",
    "plt.figure(figsize = (12,10))\n",
    "row, colums = 4, 4\n",
    "for i in range(16):  \n",
    "    plt.subplot(colums, row, i+1)\n",
    "    plt.imshow(X_shuffle.iloc[i].values.reshape(28,28), cmap=\"Greys\", interpolation='nearest')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Huruf I =  1118\n",
      "Huruf F =  1164\n"
     ]
    }
   ],
   "source": [
    "word_dict = {0:'A',1:'B',2:'C',3:'D',4:'E',5:'F',6:'G',7:'H',8:'I',9:'J',10:'K',11:'L',12:'M',13:'N',14:'O',15:'P',16:'Q',17:'R',18:'S',19:'T',20:'U',21:'V',22:'W',23:'X', 24:'Y',25:'Z',26:'A_Flip',27:'B_Flip',28:'C_Flip',29:'D_Flip',30:'E_Flip',31:'F_Flip',32:'G_Flip',33:'H_Flip',34:'I_Flip',35:'J_Flip',36:'K_Flip',37:'L_Flip',38:'M_Flip',39:'N_Flip',40:'O_Flip',41:'P_Flip',42:'Q_Flip',43:'R_Flip',44:'S_Flip',45:'T_Flip',46:'U_Flip',47:'V_Flip', 48:'W_Flip',49:'X_Flip', 50:'Y_Flip',51:'Z_Flip'}\n",
    "y_int = np.int64(y)\n",
    "count = np.zeros(52, dtype='int') #a vector of size 26 with all 0 values\n",
    "for i in y_int:\n",
    "    count[i] +=1 #total count of each alphabet\n",
    "\n",
    "alphabets = []\n",
    "for i in word_dict.values():\n",
    "    alphabets.append(i) #all alphabets\n",
    "    \n",
    "fig, ax = plt.subplots(1,1, figsize=(20,20))\n",
    "ax.barh(alphabets, count)\n",
    "\n",
    "plt.xlabel(\"Number of elements \")\n",
    "plt.ylabel(\"Alphabets\")\n",
    "plt.grid()\n",
    "plt.show()\n",
    "\n",
    "print(\"Huruf I = \", count[8])\n",
    "print(\"Huruf F = \", count[5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# splite the data\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y)\n",
    "\n",
    "# scale data\n",
    "standard_scaler = MinMaxScaler()\n",
    "standard_scaler.fit(X_train)\n",
    "\n",
    "X_train = standard_scaler.transform(X_train)\n",
    "X_test = standard_scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data after scaler\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Data after scaler\")\n",
    "X_shuffle = shuffle(X_train)\n",
    "\n",
    "plt.figure(figsize = (12,10))\n",
    "row, colums = 4, 4\n",
    "for i in range(16):\n",
    "    plt.subplot(colums, row, i+1)\n",
    "    plt.imshow(X_shuffle[i].reshape(28,28),interpolation='nearest', cmap='Greys')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train.reshape(X_train.shape[0], 28, 28, 1).astype('float32')\n",
    "X_test = X_test.reshape(X_test.shape[0], 28, 28, 1).astype('float32')\n",
    "\n",
    "y_train = np_utils.to_categorical(y_train)\n",
    "y_test = np_utils.to_categorical(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0  :  A\n",
      "1  :  B\n",
      "2  :  C\n",
      "3  :  D\n",
      "4  :  E\n",
      "5  :  F\n",
      "6  :  G\n",
      "7  :  H\n",
      "8  :  I\n",
      "9  :  J\n",
      "10  :  K\n",
      "11  :  L\n",
      "12  :  M\n",
      "13  :  N\n",
      "14  :  O\n",
      "15  :  P\n",
      "16  :  Q\n",
      "17  :  R\n",
      "18  :  S\n",
      "19  :  T\n",
      "20  :  U\n",
      "21  :  V\n",
      "22  :  W\n",
      "23  :  X\n",
      "24  :  Y\n",
      "25  :  Z\n",
      "26  :  A_Flip\n",
      "27  :  B_Flip\n",
      "28  :  C_Flip\n",
      "29  :  D_Flip\n",
      "30  :  E_Flip\n",
      "31  :  F_Flip\n",
      "32  :  G_Flip\n",
      "33  :  H_Flip\n",
      "34  :  I_Flip\n",
      "35  :  J_Flip\n",
      "36  :  K_Flip\n",
      "37  :  L_Flip\n",
      "38  :  M_Flip\n",
      "39  :  N_Flip\n",
      "40  :  O_Flip\n",
      "41  :  P_Flip\n",
      "42  :  Q_Flip\n",
      "43  :  R_Flip\n",
      "44  :  S_Flip\n",
      "45  :  T_Flip\n",
      "46  :  U_Flip\n",
      "47  :  V_Flip\n",
      "48  :  W_Flip\n",
      "49  :  X_Flip\n",
      "50  :  Y_Flip\n",
      "51  :  Z_Flip\n"
     ]
    }
   ],
   "source": [
    "for key in word_dict.keys():\n",
    "    print(key, \" : \", word_dict[key])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build, Compile, and Training Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model A (2 kali convolution)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model A\n",
    "model = Sequential()\n",
    "model.add(Conv2D(32, kernel_size = (5, 5), input_shape=(28,28,1), activation = \"relu\",  padding = \"Same\"))\n",
    "model.add(Conv2D(32, kernel_size = (5, 5), activation = \"relu\",  padding = \"Same\"))\n",
    "model.add(MaxPool2D(pool_size = (2, 2)))\n",
    "model.add(Dropout(0.50))\n",
    "\n",
    "model.add(Conv2D(64, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "model.add(Conv2D(64, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "model.add(Conv2D(64, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "model.add(MaxPool2D(pool_size = (2, 2)))\n",
    "model.add(Dropout(0.50))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128, activation = \"relu\"))\n",
    "model.add(Dropout(0.50))\n",
    "model.add(Dense(len(y.unique()), activation = \"softmax\"))\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model B (1 kali convolution, fully functional layer + 1 Dense)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model = Sequential()\n",
    "# model.add(Conv2D(32, (5, 5), input_shape=(28, 28, 1), activation='relu', padding=\"Same\"))\n",
    "# model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "# model.add(Dropout(0.3))\n",
    "# model.add(Flatten())\n",
    "# model.add(Dense(128, activation='relu'))\n",
    "# model.add(Dense(len(y.unique()), activation='softmax'))\n",
    "# model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model C (Reduced A)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# model = Sequential()\n",
    "# model.add(Conv2D(32, kernel_size = (5, 5), input_shape=(28,28,1), activation = \"relu\",  padding = \"Same\"))\n",
    "# model.add(Conv2D(32, kernel_size = (5, 5), activation = \"relu\",  padding = \"Same\"))\n",
    "# model.add(MaxPool2D(pool_size = (2, 2)))\n",
    "# model.add(Dropout(0.50))\n",
    "\n",
    "# model.add(Flatten())\n",
    "# model.add(Dense(128, activation = \"relu\"))\n",
    "# model.add(Dropout(0.50))\n",
    "# model.add(Dense(len(y.unique()), activation = \"softmax\"))\n",
    "# model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model D (?)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model = Sequential()\n",
    "# model.add(Conv2D(32, kernel_size = (5, 5), input_shape=(28,28,1), activation = \"relu\",  padding = \"Same\"))\n",
    "# model.add(Conv2D(32, kernel_size = (5, 5), activation = \"relu\",  padding = \"Same\"))\n",
    "# model.add(MaxPool2D(pool_size = (2, 2)))\n",
    "# model.add(Dropout(0.50))\n",
    "\n",
    "# model.add(Conv2D(64, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "# model.add(Conv2D(64, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "# model.add(Conv2D(64, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "# model.add(MaxPool2D(pool_size = (2, 2)))\n",
    "# model.add(Dropout(0.50))\n",
    "\n",
    "# model.add(Conv2D(128, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "# model.add(Conv2D(128, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "# model.add(Conv2D(128, kernel_size = (5, 5), activation = \"relu\", padding = \"Same\"))\n",
    "# model.add(MaxPool2D(pool_size = (2, 2)))\n",
    "# model.add(Dropout(0.50))\n",
    "\n",
    "# model.add(Flatten())\n",
    "# model.add(Dense(128, activation = \"relu\"))\n",
    "# model.add(Dropout(0.50))\n",
    "# model.add(Dense(len(y.unique()), activation = \"softmax\"))\n",
    "# model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Augmentation (Flip Horizontal and Vertical)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 100\n",
    "epochsize = 5\n",
    "\n",
    "datagen = ImageDataGenerator(rotation_range = 45,  \n",
    "                             vertical_flip = True,\n",
    "                             horizontal_flip = True)\n",
    "\n",
    "train_batches = datagen.flow(X_train, y_train, batch_size = batch_size)\n",
    "val_batches = datagen.flow(X_test, y_test, batch_size = batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-139-e674495041f8>:2: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
      "  history = model.fit_generator(train_batches,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "5581/5581 [==============================] - 148s 26ms/step - loss: 1.4989 - accuracy: 0.3863 - val_loss: 1.0073 - val_accuracy: 0.4621\n",
      "Epoch 2/5\n",
      "5581/5581 [==============================] - 146s 26ms/step - loss: 1.1569 - accuracy: 0.4411 - val_loss: 0.9622 - val_accuracy: 0.4718\n",
      "Epoch 3/5\n",
      "5581/5581 [==============================] - 152s 27ms/step - loss: 1.1138 - accuracy: 0.4512 - val_loss: 0.9491 - val_accuracy: 0.4835\n",
      "Epoch 4/5\n",
      "5581/5581 [==============================] - 155s 28ms/step - loss: 1.0908 - accuracy: 0.4579 - val_loss: 0.9239 - val_accuracy: 0.4968\n",
      "Epoch 5/5\n",
      "5581/5581 [==============================] - 148s 26ms/step - loss: 1.0744 - accuracy: 0.4646 - val_loss: 0.9155 - val_accuracy: 0.4979\n"
     ]
    }
   ],
   "source": [
    "\n",
    "#fitting the model\n",
    "history = model.fit_generator(train_batches,\n",
    "                    epochs = epochsize,\n",
    "                    validation_data = val_batches)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"import tensorflow as tf\\nmodel = tf.keras.models.load_model('model_hand.h5')\\nhistory = np.load('history.npy', allow_pickle='TRUE').item()\\nprint(history)\""
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "# Load Model and History (Model Fit)\n",
    "'''import tensorflow as tf\n",
    "model = tf.keras.models.load_model('model_hand.h5')\n",
    "history = np.load('history.npy', allow_pickle='TRUE').item()\n",
    "print(history)'''\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "scrolled": true
   },
   "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": [
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('Model loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d_7 (Conv2D)           (None, 28, 28, 32)        832       \n",
      "                                                                 \n",
      " conv2d_8 (Conv2D)           (None, 28, 28, 32)        25632     \n",
      "                                                                 \n",
      " max_pooling2d_3 (MaxPooling  (None, 14, 14, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " dropout_5 (Dropout)         (None, 14, 14, 32)        0         \n",
      "                                                                 \n",
      " conv2d_9 (Conv2D)           (None, 14, 14, 64)        51264     \n",
      "                                                                 \n",
      " conv2d_10 (Conv2D)          (None, 14, 14, 64)        102464    \n",
      "                                                                 \n",
      " conv2d_11 (Conv2D)          (None, 14, 14, 64)        102464    \n",
      "                                                                 \n",
      " max_pooling2d_4 (MaxPooling  (None, 7, 7, 64)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " dropout_6 (Dropout)         (None, 7, 7, 64)          0         \n",
      "                                                                 \n",
      " flatten_2 (Flatten)         (None, 3136)              0         \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 128)               401536    \n",
      "                                                                 \n",
      " dropout_7 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 52)                6708      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 690,900\n",
      "Trainable params: 690,900\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()\n",
    "model.save(r'model_hand.h5')\n",
    "\n",
    "np.save('history.npy', history.history)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Making Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x648 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(3,3, figsize=(8,9))\n",
    "axes = axes.flatten()\n",
    "\n",
    "for i,ax in enumerate(axes):\n",
    "    img = np.reshape(X_test[i], (28,28))\n",
    "    ax.imshow(img, cmap=\"Greys\")\n",
    "    pred = word_dict[np.argmax(y_test[i])]\n",
    "    ax.set_title(\"Prediction: \" + pred)\n",
    "    ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x648 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(3,3, figsize=(8,9))\n",
    "axes = axes.flatten()\n",
    "\n",
    "for i,ax in enumerate(axes):\n",
    "    img = np.reshape(X_test[i], (28,28)) # reshaping it for displaying\n",
    "    ax.imshow(img, cmap=\"Greys\")\n",
    "    img_final = np.reshape(img, (1,28,28,1)) # reshaping it for passing into model for prediction\n",
    "    pred = word_dict[np.argmax(model.predict(img_final))]\n",
    "    ax.set_title(\"Prediction: \"+pred)\n",
    "    ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x648 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from PIL import Image, ImageOps\n",
    "\n",
    "fig, axes = plt.subplots(3,3, figsize=(8,9))\n",
    "axes = axes.flatten()\n",
    "\n",
    "for i,ax in enumerate(axes):\n",
    "    load_name = 'cek/Sample1/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = img.convert(\"L\")\n",
    "    img = ImageOps.invert(img) # untuk white to black, black to white\n",
    "    #img = ImageOps.grayscale(img)\n",
    "    ax.imshow(img, cmap=\"Greys\")\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    ax.set_title(\"Prediction: \"+pred)\n",
    "    ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A = A_Flip\n",
      "B = B_Flip\n",
      "C = C\n",
      "D = D_Flip\n",
      "E = E\n",
      "F = E\n",
      "G = G\n",
      "H = H\n",
      "I = I\n",
      "J = J\n",
      "K = K_Flip\n",
      "L = L_Flip\n",
      "M = M_Flip\n",
      "N = N\n",
      "O = P_Flip\n",
      "P = B\n",
      "Q = Q\n",
      "R = R_Flip\n",
      "S = S_Flip\n",
      "T = J_Flip\n",
      "U = U_Flip\n",
      "V = U_Flip\n",
      "W = W\n",
      "X = X_Flip\n",
      "Y = Y_Flip\n",
      "Z = Z\n",
      "Prediksi benar = 0.8076923076923077\n"
     ]
    }
   ],
   "source": [
    "jumlahbenar = 0\n",
    "\n",
    "\n",
    "for i in range(26):\n",
    "    load_name = 'cek/Sample1/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = ImageOps.grayscale(img)\n",
    "    img = ImageOps.invert(img)\n",
    "    img = np.asarray(img)\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    if (chr(i+65) == pred or chr(i+65) + \"_Flip\" == pred):\n",
    "        jumlahbenar += 1\n",
    "    print(chr(i+65) + ' = ' + pred)\n",
    "    \n",
    "print(\"Prediksi benar = \" + str(jumlahbenar/26))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A = A_Flip\n",
      "B = B\n",
      "C = C_Flip\n",
      "D = D_Flip\n",
      "E = E\n",
      "F = R_Flip\n",
      "G = G\n",
      "H = H_Flip\n",
      "I = I_Flip\n",
      "J = J\n",
      "K = K_Flip\n",
      "L = J_Flip\n",
      "M = M_Flip\n",
      "N = N\n",
      "O = O_Flip\n",
      "P = R_Flip\n",
      "Q = Q\n",
      "R = R\n",
      "S = S_Flip\n",
      "T = K_Flip\n",
      "U = U_Flip\n",
      "V = U_Flip\n",
      "W = W\n",
      "X = X_Flip\n",
      "Y = R\n",
      "Z = Z\n",
      "Prediksi benar = 0.7692307692307693\n"
     ]
    }
   ],
   "source": [
    "jumlahbenar = 0\n",
    "\n",
    "for i in range(26):\n",
    "    load_name = 'cek/Sample2/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = ImageOps.grayscale(img)\n",
    "    img = ImageOps.invert(img)\n",
    "    img = np.asarray(img)\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    if (chr(i+65) == pred or chr(i+65) + \"_Flip\" == pred):\n",
    "        jumlahbenar += 1\n",
    "    print(chr(i+65) + ' = ' + pred)\n",
    "    \n",
    "print(\"Prediksi benar = \" + str(jumlahbenar/26))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A = A_Flip\n",
      "B = B_Flip\n",
      "C = S_Flip\n",
      "D = D_Flip\n",
      "E = B\n",
      "F = F_Flip\n",
      "G = B_Flip\n",
      "H = H\n",
      "I = U_Flip\n",
      "J = C_Flip\n",
      "K = K_Flip\n",
      "L = F_Flip\n",
      "M = W\n",
      "N = N\n",
      "O = Q\n",
      "P = B_Flip\n",
      "Q = Q\n",
      "R = R_Flip\n",
      "S = G\n",
      "T = K_Flip\n",
      "U = U_Flip\n",
      "V = N_Flip\n",
      "W = W\n",
      "X = X_Flip\n",
      "Y = R\n",
      "Z = B_Flip\n",
      "Prediksi benar = 0.46153846153846156\n"
     ]
    }
   ],
   "source": [
    "jumlahbenar = 0\n",
    "\n",
    "for i in range(26):\n",
    "    load_name = 'cek/Sample3/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = ImageOps.grayscale(img)\n",
    "    img = ImageOps.invert(img)\n",
    "    img = np.asarray(img)\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    if (chr(i+65) == pred or chr(i+65) + \"_Flip\" == pred):\n",
    "        jumlahbenar += 1\n",
    "    print(chr(i+65) + ' = ' + pred)\n",
    "    \n",
    "print(\"Prediksi benar = \" + str(jumlahbenar/26))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A = A_Flip\n",
      "B = B_Flip\n",
      "C = W\n",
      "D = D_Flip\n",
      "E = U\n",
      "F = X_Flip\n",
      "G = B\n",
      "H = H_Flip\n",
      "I = R_Flip\n",
      "J = J\n",
      "K = K_Flip\n",
      "L = Z\n",
      "M = W\n",
      "N = Z\n",
      "O = O\n",
      "P = R_Flip\n",
      "Q = Q\n",
      "R = R_Flip\n",
      "S = S_Flip\n",
      "T = K\n",
      "U = J_Flip\n",
      "V = Z_Flip\n",
      "W = E\n",
      "X = X_Flip\n",
      "Y = R_Flip\n",
      "Z = B_Flip\n",
      "Prediksi benar = 0.4230769230769231\n"
     ]
    }
   ],
   "source": [
    "jumlahbenar = 0\n",
    "\n",
    "for i in range(26):\n",
    "    load_name = 'cek/Sample4/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = ImageOps.grayscale(img)\n",
    "    img = ImageOps.invert(img)\n",
    "    img = np.asarray(img)\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    if (chr(i+65) == pred or chr(i+65) + \"_Flip\" == pred):\n",
    "        jumlahbenar += 1\n",
    "    print(chr(i+65) + ' = ' + pred)\n",
    "    \n",
    "print(\"Prediksi benar = \" + str(jumlahbenar/26))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A = A_Flip\n",
      "B = B\n",
      "C = C_Flip\n",
      "D = D_Flip\n",
      "E = B\n",
      "F = E\n",
      "G = G\n",
      "H = H\n",
      "I = I_Flip\n",
      "J = J\n",
      "K = K_Flip\n",
      "L = E_Flip\n",
      "M = M_Flip\n",
      "N = N\n",
      "O = O_Flip\n",
      "P = P_Flip\n",
      "Q = Q\n",
      "R = R_Flip\n",
      "S = S_Flip\n",
      "T = T\n",
      "U = Z_Flip\n",
      "V = V_Flip\n",
      "W = M_Flip\n",
      "X = X_Flip\n",
      "Y = R_Flip\n",
      "Z = E\n",
      "Prediksi benar = 0.7307692307692307\n"
     ]
    }
   ],
   "source": [
    "jumlahbenar = 0\n",
    "\n",
    "for i in range(26):\n",
    "    load_name = 'cek/Sample5/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = ImageOps.grayscale(img)\n",
    "    img = ImageOps.invert(img)\n",
    "    img = np.asarray(img)\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    if (chr(i+65) == pred or chr(i+65) + \"_Flip\" == pred):\n",
    "        jumlahbenar += 1\n",
    "    print(chr(i+65) + ' = ' + pred)\n",
    "    \n",
    "print(\"Prediksi benar = \" + str(jumlahbenar/26))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A = A_Flip\n",
      "B = B\n",
      "C = C\n",
      "D = D_Flip\n",
      "E = E\n",
      "F = R_Flip\n",
      "G = G\n",
      "H = H\n",
      "I = I_Flip\n",
      "J = J_Flip\n",
      "K = K_Flip\n",
      "L = L\n",
      "M = M_Flip\n",
      "N = N\n",
      "O = D_Flip\n",
      "P = R\n",
      "Q = Q\n",
      "R = R\n",
      "S = S_Flip\n",
      "T = T\n",
      "U = U_Flip\n",
      "V = U_Flip\n",
      "W = M_Flip\n",
      "X = X_Flip\n",
      "Y = K_Flip\n",
      "Z = S_Flip\n",
      "Prediksi benar = 0.7307692307692307\n"
     ]
    }
   ],
   "source": [
    "jumlahbenar = 0\n",
    "\n",
    "for i in range(26):\n",
    "    load_name = 'cek/Sample6/'+chr(i+65)+'.png'\n",
    "    img = Image.open(load_name)\n",
    "    img = ImageOps.grayscale(img)\n",
    "    img = ImageOps.invert(img)\n",
    "    img = np.asarray(img)\n",
    "    img = np.reshape(img, (1,28,28,1))\n",
    "    pred = word_dict[np.argmax(model.predict(img))]\n",
    "    if (chr(i+65) == pred or chr(i+65) + \"_Flip\" == pred):\n",
    "        jumlahbenar += 1\n",
    "    print(chr(i+65) + ' = ' + pred)\n",
    "    \n",
    "print(\"Prediksi benar = \" + str(jumlahbenar/26))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}