

In [1]:

    
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt



In [2]:

    
%matplotlib inline



In [3]:

    
a=tf.random_normal([2,20])



In [4]:

    
sess=tf.Session()



In [5]:

    
out=sess.run(a)



In [6]:

    
print(out)









    



[[-1.01132369 -0.03734883 -1.50488603  0.05773655  0.66317093 -1.22423506
  -1.81542122  1.57534635 -0.14326067  2.08635116 -0.79781163 -1.12896252
   1.26442504 -0.26508138 -0.05267891  0.99382091  0.29550475  0.50181234
   1.28935134  1.03054881]
 [-0.79155535  0.21548827 -0.07632679 -0.20849626  0.64689469 -0.98350734
   1.59773135  2.77432394 -0.90824676 -1.14271474 -0.55331206 -0.81772059
  -0.02598415  0.43808228 -0.93310952 -0.33111054  1.91047168  0.71959078
  -0.75361609  1.29549837]]



In [7]:

    
x,y=out



In [8]:

    
plt.scatter(x,y)









    Out[8]:





<matplotlib.collections.PathCollection at 0x7fbc48beb9b0>



In [9]:

    
a=tf.constant(5,name='input_a')
b=tf.constant(3,name='input_b')
c=tf.multiply(a,b,name='mul_c')
d=tf.add(a,b,name='add_d')
e=tf.add(c,d,name='add_e')



In [10]:

    
sess=tf.Session()



In [11]:

    
sess.run(e)









    Out[11]:





23



In [16]:

    
writer=tf.summary.FileWriter('my_graph',sess.graph)

The above command will output graph to my_graph directory Then run the command '$ tensorboard --logdir="my_graph"' type http://localhost:6006 in the browser



In [17]:

    
writer.close()



In [18]:

    
sess.close()



In [3]:

    
a=tf.constant([5,3],name='input_a')



In [6]:

    
b=tf.reduce_prod(a,name='prod_b')



In [7]:

    
c=tf.reduce_sum(a,name='sum_c')



In [8]:

    
d=tf.add(b,c,name='add_d')



In [9]:

    
sess=tf.Session()



In [10]:

    
sess.run(d)









    Out[10]:





23

Tensors, as mentioned before, are simply the n-dimensional abstractionof matrices. So a 1-D tensor would be equivalent to a vector, a 2-Dtensor is a matrix, and above that you can just say “N-D tensor”.

Comparing to last example, this one has only one input which is a vector, so simplify the graph.And also reduce to only one dependency to check.



In [12]:

    
help(tf.placeholder)









    



Help on function placeholder in module tensorflow.python.ops.array_ops:

placeholder(dtype, shape=None, name=None)
    Inserts a placeholder for a tensor that will be always fed.
    
    **Important**: This tensor will produce an error if evaluated. Its value must
    be fed using the `feed_dict` optional argument to `Session.run()`,
    `Tensor.eval()`, or `Operation.run()`.
    
    For example:
    
    ```python
    x = tf.placeholder(tf.float32, shape=(1024, 1024))
    y = tf.matmul(x, x)
    
    with tf.Session() as sess:
      print(sess.run(y))  # ERROR: will fail because x was not fed.
    
      rand_array = np.random.rand(1024, 1024)
      print(sess.run(y, feed_dict={x: rand_array}))  # Will succeed.
    ```
    
    Args:
      dtype: The type of elements in the tensor to be fed.
      shape: The shape of the tensor to be fed (optional). If the shape is not
        specified, you can feed a tensor of any shape.
      name: A name for the operation (optional).
    
    Returns:
      A `Tensor` that may be used as a handle for feeding a value, but not
      evaluated directly.



In [13]:

    
a=tf.placeholder(tf.int32,shape=(2),name='input_a')



In [14]:

    
b=tf.reduce_prod(a,name='prod_b')



In [15]:

    
c=tf.mul(b,3,name='mul_c')



In [16]:

    
sess=tf.Session()



In [19]:

    
input_dict={a:np.array([3,5],dtype=np.int32)}



In [20]:

    
sess.run(c,feed_dict=input_dict)









    Out[20]:





45

This way create a placeholder which hold variable values