def mixed_mnist(count):
total = count*10
valid = int(np.floor(count/10))
rawset = mnist_data()
arranged_data = np.zeros((1,len(rawset[0][0])))
arranged_target = np.zeros((1,1))
for i in range(10):
arranged_data = np.append(arranged_data,rawset[0][500*i:(500*i)+count],axis=0)
arranged_target = np.append(arranged_target,np.ones((count,),dtype=int)*i)
for i in range(10): #validation
arranged_data = np.append(arranged_data,rawset[0][500*(i+1)-valid:500*(i+1)],axis=0)
arranged_target = np.append(arranged_target,np.ones((valid,),dtype=int)*i)
arranged_data = np.delete(arranged_data,0,axis=0)
arranged_target = np.delete(arranged_target,0,axis=0)
return [arranged_data,arranged_target,valid*10]
def load_train_test(test_size=0.33, random_state=2):
X, y = mnist_data()
X_train, X_test, y_train, y_test = train_test_split(
X, y, test_size=test_size, random_state=random_state, stratify=y)
X_train = X_train.reshape(X_train.shape[0], 1, 28, 28).astype('float32')
X_test = X_test.reshape(X_test.shape[0], 1, 28, 28).astype('float32')
X_train = X_train / 255
X_test = X_test / 255
# one hot encode outputs
y_train = np_utils.to_categorical(y_train)
y_test = np_utils.to_categorical(y_test)
return X_train, X_test, y_train, y_test
def test_import_mnist_data():
X, y = mnist_data()
assert (X.shape[0] == 5000)
assert (X.shape[1] == 784)
assert (y.shape[0] == 5000)
# mnist.py
from mlxtend.data import mnist_data
from mlxtend.classifier import MultiLayerPerceptron as MLP
from mlxtend.preprocessing import shuffle_arrays_unison
X, y = mnist_data()
X, y = shuffle_arrays_unison((X, y), random_seed=1)
X_train, y_train = X[:500], y[:500]
X_test, y_test = X[500:], y[500:]
import matplotlib.pyplot as plt
def plot_digit(X, y, idx):
img = X[idx].reshape(28, 28)
plt.imshow(img, cmap='Greys', interpolation='nearest')
plt.title('true label: %d' % y[idx])
plt.show()
plot_digit(X, y, 3500)
from mlxtend.preprocessing import standardize
X_train_std, params = standardize(X_train,
columns=range(X_train.shape[1]),
return_params=True)
X_test_std = standardize(X_test, columns=range(X_test.shape[1]), params=params)
nn1 = MLP(hidden_layers=[150],
l2=0.00,
l1=0.0,
epochs=100,
def test_import_mnist_data():
X, y = mnist_data()
assert(X.shape[0] == 5000)
assert(X.shape[1] == 784)
assert(y.shape[0] == 5000)
from mlxtend.data import mnist_data
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import math
from math import ceil
import tensorflow as tf
from sklearn import model_selection
# load in the data
X_orig, y_orig = mnist_data()
# print(X[1])
image_width = image_height = int(math.sqrt(len(X_orig[0])))
# reshape X
X_orig = np.array(X_orig).astype(float)
X = X_orig.reshape([X_orig.shape[0], image_height, image_height, 1])
# form train & test set
X_train, X_test, Y_train, Y_test = model_selection.train_test_split(X, y_orig, test_size=0.02, random_state=17)
# change y to one_hot vector
def convert_to_onehot(y_vec):
"""
A function that reshapes the y vector (1,2,5...) into one hot tensor[[0,1,0,0,0,0,0...], [0,0,1,0,0,0,0...]]
:param y_vec
:return: one_hot tensor
"""
classes = len(set(y_vec))
classes_tf = tf.constant(classes)
__author__ = 'BENGEOS-PC'
from mlxtend.data import mnist_data
from matplotlib import pyplot as plt
def plot_digit(X,y,idx):
img = X[idx].reshape(28,28)
plt.imshow(img,cmap='Greys',interpolation='nearest')
plt.title('True Lable: %'% y[idx])
plt.show()
X,y = mnist_data()
plot_digit(X,y,4)
