def prepare_data(self):
(x_train, y_train), (x_eval, y_eval) = boston_housing.load_data()
ss = StandardScaler()
ss.fit(x_train)
x_train, x_eval = ss.transform(x_train), ss.transform(x_eval)
train_data, eval_data = (x_train, y_train), (x_eval, y_eval)
return train_data, eval_data
def prepare_data(self):
from tensorflow.python.keras.datasets import boston_housing
from sklearn.preprocessing import StandardScaler
(x_train, y_train), (x_eval, y_eval) = boston_housing.load_data()
ss = StandardScaler()
ss.fit(x_train)
x_eval = ss.transform(x_eval)
return x_eval, y_eval
def __init__(self):
(x_train, y_train), (x_test, y_test) = boston_housing.load_data()
mean = x_train.mean(axis=0)
x_train -= mean
std = x_train.std(axis=0)
x_train /= std
x_test -= mean
x_test /= std
super().__init__(x_train, x_test, y_train, y_test, (x_train.shape[1:]),
1, 'boston_housing')
def load_data():
'''Loads and normalizes the boston housing data'''
(train_data, train_targets), (test_data, test_targets) =\
boston_housing.load_data()
mean = train_data.mean(axis=0)
train_data -= mean
std = train_data.std(axis=0)
train_data /= std
test_data -= mean
test_data /= std
return (train_data, train_targets), (test_data, test_targets)
from tensorflow.contrib.eager.python import tfe
# enable eager mode
tf.enable_eager_execution()
tf.set_random_seed(0)
np.random.seed(0)
if not os.path.exists('weights/'):
os.makedirs('weights/')
# constants
batch_size = 128
epochs = 25
# dataset loading
(x_train, y_train), (x_test, y_test) = boston_housing.load_data()
# normalization of dataset
mean = x_train.mean(axis=0)
std = x_train.std(axis=0)
x_train = (x_train - mean) / (std + 1e-8)
x_test = (x_test - mean) / (std + 1e-8)
print('x train', x_train.shape, x_train.mean(), x_train.std())
print('y train', y_train.shape, y_train.mean(), y_train.std())
print('x test', x_test.shape, x_test.mean(), x_test.std())
print('y test', y_test.shape, y_test.mean(), y_test.std())
# model definition (canonical way)
def prepare_data(self):
from tensorflow.python.keras.datasets import boston_housing
(x_train, y_train), (x_eval, y_eval) = boston_housing.load_data()
x_train, x_eval = standardize(x_train, x_eval)
train_data, eval_data = (x_train, y_train), (x_eval, y_eval)
return train_data, eval_data
from tensorflow.python.keras.datasets import boston_housing (train_data,train_targets),(test_data,test_targets) =boston_housing.load_data() print(train_data.shape)
#from google.colab import drive
#drive.mount('/content/drive')
#/content/drive/My Drive/ANN Mahesh Anand/
# ### Collect Data
# In[1]:
from tensorflow.python.keras.datasets import boston_housing
#Load data
(features, actual_prices),_ = boston_housing.load_data(test_split=0)
# In[2]:
print('Number of examples: ', features.shape[0])
print('Number of features for each example: ', features.shape[1])
print('Shape of actual prices data: ', actual_prices.shape)
# # Building the graph
# Define input data placeholders
# In[6]:
from tensorflow.python.keras.datasets import boston_housing
data_path = "D:\\data\\boston_housing.npz"
(train_datas,train_targets),(test_datas,test_targets) = boston_housing.load_data(path=data_path)
print(train_datas.shape)
#数据标准化
#每列的平均值
mean= train_datas.mean(axis=0)
#减去平均值
train_datas -= mean
#每列的
std = train_datas.std(axis=0)
train_datas /= std
test_datas -=mean
test_datas /=std
#构建网络
from tensorflow.python.keras import layers,models
def build_model():
model = models.Sequential()
model.add(layers.Dense(64,activation='relu',input_shape=(train_datas.shape[1],)))
model.add(layers.Dense(64,activation='relu'))
model.add(layers.Dense(1))
model.compile(
optimizer='rmsprop',
loss= 'mse',
metrics=['mae']
)
return model
get_session()
tfe.enable_eager_execution()
tfe.executing_eagerly() # => True
tf.set_random_seed(0)
np.random.seed(0)
if not os.path.exists('weights/'):
os.makedirs('weights/')
# 2. parameters
batch_size = 128
epochs = 100
# 3. train data
(x_train, y_train), (x_test, y_test) = boston_housing.load_data(test_split=0.1)
mean = x_train.mean(axis=0)
std = x_train.std(axis=0)
x_train = (x_train - mean) / (std + 1e-8)
x_test = (x_test - std) / (std + 1e-8)
print('x train', x_train.shape, x_train.mean(), x_train.std())
print('y train', y_train.shape, y_train.mean(), y_train.std())
print('x test', x_test.shape, x_test.mean(), x_test.std())
print('y test', y_test.shape, y_test.mean(), y_test.std())
# 4. model (linear regression)
def build_model(input_shape=None):
def prepare_data(self):
(x_train, y_train), (x_test, y_test) = boston_housing.load_data()
ss = StandardScaler()
ss.fit(x_train)
x_test = ss.transform(x_test)
return x_test
def prepare_data(self):
from tensorflow.python.keras.datasets import boston_housing
(x_train, y_train), (x_test, y_test) = boston_housing.load_data()
_, x_test = standardize(x_train, x_test)
return x_test
'''
@assignment: Lab 6, Exercise 3
@student: Sarah Whitten
@date: March 14, 2020
'''
import numpy as np
from tensorflow.python.keras.datasets import boston_housing
# load data
(train_images, train_labels), (test_images,
test_labels) = boston_housing.load_data()
# print number of training and testing examples
# from the class example guide
def print_structures():
print(
'training images \
\n\tcount: {} \
\n\tdimensions: {} \
\n\tshape: {} \
\n\tdata type: {}\n\n'.format(len(train_images), train_images.ndim,
train_images.shape,
train_images.dtype),
'testing images \
\n\tcount: {} \
\n\tdimensions: {} \
\n\tshape: {} \
\n\tdata type: {} \
\n\tvalues: {}\n'.format(len(test_labels), train_labels.ndim,