TensorFlow Basics

In [1]:

    

import tensorflow as tf


    

    




c:\programdata\anaconda3\lib\site-packages\h5py\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters

In [2]:

    

print(tf.__version__)


    

    




1.7.0

In [3]:

    

hello = tf.constant('Hello')


    

In [4]:

    

type(hello)


    

    
Out[4]:





tensorflow.python.framework.ops.Tensor

In [5]:

    

world = tf.constant('World')


    

In [6]:

    

result = hello + world


    

In [7]:

    

result


    

    
Out[7]:





<tf.Tensor 'add:0' shape=() dtype=string>

In [8]:

    

type(result)


    

    
Out[8]:





tensorflow.python.framework.ops.Tensor

In [9]:

    

with tf.Session() as sess:
    result = sess.run(hello + world)


    

In [10]:

    

result


    

    
Out[10]:





b'HelloWorld'

In [11]:

    

tensor_1 = tf.constant(1)
tensor_2 = tf.constant(2)


    

In [12]:

    

type(tensor_1)


    

    
Out[12]:





tensorflow.python.framework.ops.Tensor

In [13]:

    

tensor_1 + tensor_2


    

    
Out[13]:





<tf.Tensor 'add_2:0' shape=() dtype=int32>

In [14]:

    

sess = tf.Session()


    

In [15]:

    

sess.run(tensor_1 + tensor_2)


    

    
Out[15]:





3

In [16]:

    

sess.close()


    

In [17]:

    

# Constant
const = tf.constant(10)


    

In [18]:

    

# Create a tensor of a specific dimension
# with a given value
fill_mat = tf.fill(dims = (4, 4), value = 10)


    

In [19]:

    

# Generate a tensor of zeros
myzeros = tf.zeros(shape = (4, 4))


    

In [20]:

    

# Generate a tensor of ones
myones = tf.ones(shape = (4, 4))


    

In [21]:

    

# Generate a tensor of random numbers
# Sampled from the normal distribution
# With given mean and standard deviation
myrandn = tf.random_normal(shape = (4, 4), 
                           mean = 0,
                           stddev = 0.5)


    

In [22]:

    

# Generate a tensor of random numbers
# Sampled from the uniform distribution
# With given min and max values
myrandu = tf.random_uniform(shape = (4, 4), 
                            minval = 0,
                            maxval = 1)


    

In [23]:

    

my_ops = [const, fill_mat, myzeros, myones, myrandn, myrandu]


    

In [24]:

    

# Using an Interactive Session 
sess = tf.InteractiveSession()


    

In [25]:

    

# For interactive session 
# Some functions allow you to run .eval() on them
for op in my_ops:
    print(op.eval())
    print('\n')


    

    




10


[[10 10 10 10]
 [10 10 10 10]
 [10 10 10 10]
 [10 10 10 10]]


[[0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]]


[[1. 1. 1. 1.]
 [1. 1. 1. 1.]
 [1. 1. 1. 1.]
 [1. 1. 1. 1.]]


[[ 0.38301593  0.51866984  0.20885724  0.18511339]
 [ 0.14179526 -0.4602855   0.43054888  0.6652547 ]
 [-0.12187407  0.08375393  0.6937749   0.21612667]
 [-0.07789148  0.35029742 -0.09007152 -0.16522272]]


[[0.7571168  0.62784684 0.73356867 0.9508685 ]
 [0.5720631  0.07241786 0.44787788 0.01825118]
 [0.23339236 0.7124164  0.76858854 0.82055473]
 [0.01049113 0.6794585  0.76604354 0.47718096]]

In [26]:

    

a = tf.constant([ [1, 2],
                  [3, 4] ])


    

In [27]:

    

# Shape: 2 x 2
a.get_shape()


    

    
Out[27]:





TensorShape([Dimension(2), Dimension(2)])

In [28]:

    

# Shape: 2 x 1
b = tf.constant([[10], [100]])


    

In [29]:

    

b.get_shape()


    

    
Out[29]:





TensorShape([Dimension(2), Dimension(1)])

In [30]:

    

result = tf.matmul(a,b)


    

In [31]:

    

result.eval()


    

    
Out[31]:





array([[210],
       [430]])