深度学习之tensorflow框架（上）

 1 import tensorflow as tf
 2 import os
 3 os.environ['TF_CPP_MIN_LOG_LEVEL']='2'
 4 def tensorflow_demo():
 5 
 6     #原生python加法运算
 7     a = 2;
 8     b=3;
 9     c=a+b;
10     print("普通加法运算的结果:\n",c);
11     #tensorflow实现加法运算
12     a_t=tf.constant(2)
13     b_t=tf.constant(3)
14     c_t=a_t+b_t
15     print("tensorflow的加法运算结果:\n",c_t)
16     #开启会话
17     with tf.compat.v1.Session() as sess:
18         c_t_value = sess.run(c_t)
19         print("c_t_value:\n", c_t_value)
20     return None;
21 
22 def graph_demo():
23     """
24     图的演示
25     :return:
26     """
27     #TensorFlow实现加法运算
28     a_t = tf.constant(2)
29     b_t = tf.constant(3)
30     c_t = a_t + b_t
31     print("a_t:\n",a_t)
32     print("b_t:\n", b_t)
33     print("TensorFlow加法运算的结果：\n",c_t)
34     #查看默认图
35     #方法1:调用方法
36     default_g = tf.compat.v1.get_default_graph()
37     print("defaut_g:\n", default_g)
38     #方法2:查看属性
39     print("a_t的图属性:\n",a_t.graph)
40     print("c_t的图属性:\n", a_t.graph)
41 
42     # 自定义图
43     new_g = tf.Graph()
44     # 在自己的图中定义数据和操作
45     with new_g.as_default():
46         a_new = tf.constant(20)
47         b_new = tf.constant(30)
48         c_new = a_new + b_new
49         print("c_new:\n", c_new)
50         print("a_new的图属性:\n", a_new.graph)
51         print("c_new的图属性:\n", c_new.graph)
52 
53     # 开启会话
54     with tf.compat.v1.Session() as sess:
55         c_t_value = sess.run(c_t)
56         print("c_t_value:\n", c_t_value)
57         print("sess的图属性:\n", sess.graph)
58         # 将图写入本地生成events文件
59         tf.compat.v1.summary.FileWriter("./tmp/summary",graph=sess.graph)
60 
61 
62     with tf.compat.v1.Session(graph=new_g) as new_sess:
63         c_new_value = new_sess.run(c_new)
64         print("c_new_value:\n", c_new_value)
65         print("new_sess的图属性:\n",new_sess.graph)
66 
67 
68 
69     return None
70 if __name__ == "__main__":
71     #代码1：TensorFlow的基本结构
72     #tensorflow_demo()
73     #代码2:图的演示
74     graph_demo()

代码不同函数里，有图的演示和Tensorflow基本结构