def reshape_bw(data):
sh = data.shape
return data.reshape((sh[0], sh[1], sh[2], 1))
def reshape_flatten(data):
return data.reshape([data.shape[0], data.shape[1] * data.shape[2]])
if __name__ == "__main__":
f1 = open('./results_naive.txt', 'w+')
# for ds_name in emnist.list_datasets():
f1.write('-' * 20 + '\nRunning on the dataset ' +
emnist.list_datasets()[0] + '\n' + '-' * 20 + '\n')
print('-' * 20 + '\nRunning on the dataset ' + emnist.list_datasets()[0] +
'\n' + '-' * 20)
(train_images,
train_labels), (test_images,
test_labels) = load_data(emnist.list_datasets()[0])
train_images = reshape_bw(train_images)
test_images = reshape_bw(test_images)
train_images, test_images = train_images / 255.0, test_images / 255.0
# CNN = cnn.ConvNN()
# CNN.make_model(train_images, train_labels)
# CNN.fit(test_images, test_labels)
# CNN.plot_all(test_images, test_labels)
from keras.layers.convolutional import MaxPooling2D
from keras import backend as K
from tensorflow.keras.datasets import mnist
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix
#ignore warning messages
import warnings
warnings.filterwarnings('ignore')
sns.set()
# pip install emnist
# Import Dataset(s)
from emnist import list_datasets
list_datasets()
from emnist import extract_training_samples
images_train, labels_train = extract_training_samples('letters')
from emnist import extract_test_samples
images_test, labels_test = extract_test_samples('letters')
# Flatten Data
dims = images_train.shape[1] * images_train.shape[2]
X_train = images_train.reshape(images_train.shape[0], dims)
X_test = images_test.reshape(images_test.shape[0], dims)
# Rescale to 0 -> 1 by dividing by max pixel value (255)
X_train = X_train.astype('float32') / 255
X_test = X_test.astype('float32') / 255
import matplotlib.pyplot as plt
np.random.seed(1671) # for reproducibility
# network and training
NB_EPOCH = 100
BATCH_SIZE = 256
VERBOSE = 2
NB_CLASSES = 26 # number of outputs = number of digits
OPTIMIZER = Adam() # optimizer, explainedin this chapter
N_HIDDEN = 512
VALIDATION_SPLIT=0.2 # how much TRAIN is reserved for VALIDATION
DROPOUT = 0.20
print(list_datasets())
X_train, y_train = extract_training_samples('letters')
print("train shape: ", X_train.shape)
print("train labels: ",y_train.shape)
X_test, y_test = extract_test_samples('letters')
print("test shape: ",X_test.shape)
print("test labels: ",y_test.shape)
#for indexing from 0
y_train = y_train-1
y_test = y_test-1
#X_train is 124800 rows of 28x28 values --> reshaped in 124800 x 784
#X_train is 20800 rows of 28x28 values --> reshaped in 20800 x 784
RESHAPED = X_train.shape[1]*X_train.shape[2]
