Understanding Constant and Variable in tensorflow 2.x

Understanding Constant and Variable in TensorFlow 2.x

CONSTANT

tf.constant is immutable value it means these value will not change in future when code is run for example you defined a bias that is not changing throughout the code execution

Note : by default constant take dtype as  int

Let's see how to declare constant in tensorflow

a = tf.constant(value, dtype=None, shape=None)

dtype is your data type you can define it by tf.int32 , tf.float32

shape is your data /matrix shape

Defining 1D array in tensorflow

import tensorflow as tf

a = tf.constant([1,2,3,4] , dtype = tf.int32)

print(a)  

#print(a) give output like this 
tf.Tensor([1 2 3 4], shape=(4,), dtype=int32)

print(a.numpy())

#print(a.numpy()) give output like this 

[1,2,3,4]

Defining 2D array in Tensorflow

import tensorflow as tf

a = tf.constant([[1,2,3],[4,5,6]],dtype = tf.float32)

print(a.numpy())

output =

[[1. 2. 3.]

 [4. 5. 6.]]

notice there are decimal after every number in output in this case the reason is that we define the variable data type as float

Defining Shape in Tensorflow

import tensorflow as tf

a = tf.constant(0,shape=(2,3))

print(a.numpy())

output:

[[0 0 0]

 [0 0 0]]

if you want to know dtype or shape of your constant just type

print(a.shape)  #note here  a is my constant name

print(a.dtype)

VARIABLE

Variable is mutable in tensorflow it means its value can be change in future while executing the code. 

Example : weights defining in neural network they will change as our code execute till they didn't get optimum value

Defining 1D array in tensorflow

import tensorflow as tf

a = tf.Variable([1,2,3,4] , dtype = tf.int32)

print(a)  

#print(a) give output like this 
tf.Tensor([1 2 3 4], shape=(4,), dtype=int32)

print(a.numpy())

#print(a.numpy()) give output like this 

[1,2,3,4]

Defining 2D array in Tensorflow

import tensorflow as tf

a = tf.Variable([[1,2,3],[4,5,6]],dtype = tf.float32)

print(a.numpy())

output =

[[1. 2. 3.]

 [4. 5. 6.]]

notice there are decimal after every number in output in this case the reason is that we define the variable data type as float

Defining Shape in Tensorflow

import tensorflow as tf

a = tf.Variable(0,shape=(2,3))

print(a.numpy())

output:

[[0 0 0]

 [0 0 0]]

if you want to know dtype or shape of your constant just type

print(a.shape)  #note here  a is my Variable name

print(a.dtype)

we are going to learn more about variable modification addition in later on articles

>>Math operation in tensorfow 2.x