from mnist_cnn import DeepCNN
import create_data
from keras.utils import to_categorical
from sklearn.model_selection import train_test_split
import numpy as np
print("Preparing Data...")
num_classes = 10
input_shape = (28, 28, 1)
(x_train, y_train), (x_test, y_test) = create_data.load_data()
x_train, x_validation, y_train, y_validation = train_test_split(x_train,
y_train,
test_size=0.2)
x_train = x_train.reshape(x_train.shape[0], 28, 28, 1)
x_validation = x_validation.reshape(x_validation.shape[0], 28, 28, 1)
x_test = x_test.reshape(x_test.shape[0], 28, 28, 1)
x_train = x_train.astype('float32')
x_validation = x_validation.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_validation /= 255
x_test /= 255
y_train = to_categorical(y_train, num_classes)
y_validation = to_categorical(y_validation, num_classes)
y_test = to_categorical(y_test, num_classes)
maxi=maxarea):
full_array[k] = 1
k += 1
# conmat[get_score(true, 13390*16), get_score_unet(pred, bias, maxi=maxarea)] +=1
# print(dirs_staexp, '\n', conmat)
# tissue = np.count_nonzero(val[:,0]==imnum)*128*128
# print("undecided %: ", np.count_nonzero(predmap==3)/tissue)
# a.append(np.count_nonzero(predmap==3)/tissue)
return full_array, a
if __name__ == '__main__':
tic = time.perf_counter()
histoImages, masks = load_data()
# network
patchsize = 128
reso = 4
features = 48
blocks = 5
bias = 13390
# loading
[mean, std] = np.load('Processed/ps' + str(patchsize) + 'reso' +
str(reso) + 'select/meanstd.npy')
train = np.load('Processed/ps' + str(patchsize) + 'reso' + str(reso) +
'select/traindata.npy')
val = np.load('Processed/ps' + str(patchsize) + 'reso' + str(reso) +
'select/testdata.npy')
# catsvdogs.py
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import MaxPooling2D
from keras.layers.convolutional import Conv2D
from keras.utils import np_utils
from create_data import load_data, load_test_data
#load_data()
(X_train, y_train, X_test, y_test) = load_data()
# Plotting a sample image with label
#plt.imshow(X_train[0])
#plt.title(y_train[0])
#plt.show()
# Transforming dataset from (n, width, height) to (n, depth, width, height)
X_train = X_train.reshape(X_train.shape[0], 50, 50, 1)
X_test = X_test.reshape(X_test.shape[0], 50, 50, 1)
# The second part of the processing is to convert data type to float32 and normalise data values
# to the range of (0 - 1) instead of being (0 - 255)
X_train = X_train.astype('float32')
X_test = X_test.astype('float32')
X_train /= 255
X_test /= 255