from tensorflow.keras.datasets import mnist

(training_dataset_x, training_dataset_y), (test_dataset_x, test_dataset_y) = mnist.load_data()

import matplotlib.pyplot as plt

figure = plt.gcf()
figure.set_size_inches(10, 10)
for i in range(1, 10):
    plt.subplot(3, 3, i)
    axis = plt.gca()
    axis.set_title(str(training_dataset_y[i]))
    plt.imshow(training_dataset_x[i], cmap='gray')

plt.show()

training_dataset_x = training_dataset_x / 255
test_dataset_x = test_dataset_x / 255

from tensorflow.keras.utils import to_categorical

training_dataset_y = to_categorical(training_dataset_y)
test_dataset_y = to_categorical(test_dataset_y)

from tensorflow.keras.models import Model
from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, Flatten, Dense

inp = Input(shape=(28, 28, 1), name='Input')
x = Conv2D(32, kernel_size=(3, 3), padding='same', activation='relu', name='Convolution-1')(inp)
x = MaxPooling2D(name='Pooling-1', padding='same')(x)
x = Conv2D(64, kernel_size=(3, 3), padding='same', activation='relu', name='Convolution-2')(x)
x = MaxPooling2D(name='Pooling-2', padding='same')(x)
x = Flatten(name='Flatten')(x)
x = Dense(128, activation='relu', name='Dense-1')(x)
x = Dense(128, activation='relu', name='Dense-2')(x)
x = Dense(10, activation='softmax', name='Output')(x)

model = Model(inputs=[inp], outputs=[x], name='CONV-MNITS')
model.summary()

model.compile(optimizer='rmsprop', loss=['categorical_crossentropy'], metrics=['categorical_accuracy'])
hist = model.fit([training_dataset_x.reshape(-1, 28, 28, 1)], [training_dataset_y], batch_size=32, epochs=5, validation_split=0.2)

import matplotlib.pyplot as plt

figure = plt.gcf()
figure.set_size_inches((15, 5))
plt.title('Loss - Epoch Graphics')
plt.xlabel('Epoch')
plt.ylabel('Loss')
plt.plot(range(1, len(hist.history['loss']) + 1), hist.history['loss'])
plt.plot(range(1, len(hist.history['val_loss']) + 1), hist.history['val_loss'])
plt.legend(['Loss', 'Validation Loss'])
plt.show()

figure = plt.gcf()
figure.set_size_inches((15, 5))
plt.title('Categorical Accuracy - Epoch Graphics')
plt.xlabel('Epoch')
plt.ylabel('Categorical Accuracy')
plt.plot(range(1, len(hist.history['categorical_accuracy']) + 1), hist.history['categorical_accuracy'])
plt.plot(range(1, len(hist.history['val_categorical_accuracy']) + 1), hist.history['val_categorical_accuracy'])
plt.legend(['Categorical Accuracy', 'Validation Categorical Accuracy'])
plt.show()

eval_result = model.evaluate([test_dataset_x.reshape(-1, 28, 28, 1)], [test_dataset_y])
for i in range(len(eval_result)):
    print(f'{model.metrics_names[i]} ---> {eval_result[i]}')

import numpy as np
import glob

for path in glob.glob('test-images/*.jpg'):
    img_data = plt.imread(path)
    gray_img_data = np.average(img_data, weights=(0.3, 0.59, 0.11), axis=2)
    gray_img_data = gray_img_data / 255

    predict_result = model.predict([gray_img_data.reshape(-1, 28, 28, 1)])
    number = np.argmax(predict_result[0])
    plt.title(f'Predicted: {number}', fontsize=16)
    plt.imshow(gray_img_data, cmap='gray')
    plt.show()