1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
| '''
A linear regression learning algorithm example using TensorFlow library.
Author: Aymeric Damien
Project: https://github.com/aymericdamien/TensorFlow-Examples/
'''
#%%
from __future__ import print_function
import tensorflow as tf
import numpy
import matplotlib.pyplot as plt
rng = numpy.random
# Parameters
learning_rate = 0.02
training_epochs = 1500
display_step = 100
# Training Data
# asarray：转换为ndarray对象
train_X = numpy.asarray([3.3,4.4,5.5,6.71,6.93,4.168,9.779,6.182,7.59,2.167,
7.042,10.791,5.313,7.997,5.654,9.27,3.1])
train_Y = numpy.asarray([1.7,2.76,2.09,3.19,1.694,1.573,3.366,2.596,2.53,1.221,
2.827,3.465,1.65,2.904,2.42,2.94,1.3])
n_samples = train_X.shape[0]
# tf Graph Input
# 声明float32类型，未定义形状（shape）
X = tf.placeholder("float")
Y = tf.placeholder("float")
#print(X)
# Set model weights
# 创建变量W和b，作为线性方程的参数，并初始化为float随机数
W = tf.Variable(rng.randn(), name="weight", trainable=True)
b = tf.Variable(rng.randn(), name="bias")
# Construct a linear model
# 线性模型 pred = WX + b
pred = tf.add(tf.multiply(X, W), b)
# Mean squared error
# 均方误差 cost = E(pred - Y)^2 表示梯度下降中的代价函数，值越小表示越拟合数据
# 1/2系数使得平方求梯度后常数系数为1，方便计算，系数对结果不影响
cost = tf.reduce_sum(tf.pow(pred-Y, 2))/(2*n_samples)
# Gradient descent
# Note, minimize() knows to modify W and b because Variable objects are trainable=True by default
# 使用梯度下降算法模型优化器。minimize(cost)包括compute_gradients(cost)和apply_gradients()。
# compute_gradients(cost)计算cost的梯度，默认使用GraphKeys.TRAINABLE_VARIABLES，所以包括了变量W和b
# apply_gradients()应用梯度到变量列表，变量W和b为trainable，在这一步更新变量W和b的值
optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)
# Initialize the variables (i.e. assign their default value)
# 初始化变量的操作
init = tf.global_variables_initializer()
# Start training
with tf.Session() as sess:
# Run the initializer
# 初始化，初始化变量
sess.run(init)
# Fit all training data
for epoch in range(training_epochs): #训练次数
for (x, y) in zip(train_X, train_Y): #将X,Y打包成(xi,yi)对的形式
sess.run(optimizer, feed_dict={X: x, Y: y}) #将每组(xi,yi)代入优化器计算
# 用下面这句替代上面一层循环，得到接近的结果，但速度明显提升，为什么不这样用，存疑。
# sess.run(optimizer, feed_dict={X: train_X, Y:train_Y})
# Display logs per epoch step
# 每display_step次输出状态
if (epoch+1) % display_step == 0:
c = sess.run(cost, feed_dict={X: train_X, Y:train_Y})
print("Epoch:", '%04d' % (epoch+1), "cost=", "{:.9f}".format(c), \
"W=", sess.run(W), "b=", sess.run(b))
print("Optimization Finished!")
training_cost = sess.run(cost, feed_dict={X: train_X, Y: train_Y})
print("Training cost=", training_cost, "W=", sess.run(W), "b=", sess.run(b), '\n')
# Graphic display
# 图形化等
plt.plot(train_X, train_Y, 'ro', label='Original data')
plt.plot(train_X, sess.run(W) * train_X + sess.run(b), label='Fitted line')
plt.legend()
plt.show()
# Testing example, as requested (Issue #2)
test_X = numpy.asarray([6.83, 4.668, 8.9, 7.91, 5.7, 8.7, 3.1, 2.1])
test_Y = numpy.asarray([1.84, 2.273, 3.2, 2.831, 2.92, 3.24, 1.35, 1.03])
print("Testing... (Mean square loss Comparison)")
testing_cost = sess.run(
tf.reduce_sum(tf.pow(pred - Y, 2)) / (2 * test_X.shape[0]),
feed_dict={X: test_X, Y: test_Y}) # same function as cost above
print("Testing cost=", testing_cost)
print("Absolute mean square loss difference:", abs(
training_cost - testing_cost))
plt.plot(test_X, test_Y, 'bo', label='Testing data')
plt.plot(train_X, sess.run(W) * train_X + sess.run(b), label='Fitted line')
plt.legend()
plt.show()
|