What is the shared variable in theano?

Recipe Objective - What is a shared variable in theano?

The shared function creates what is known as shared variables. These variables are a mix of symbolic and non-symbolic variables with a value that can be shared across several functions. It's referred to as a shared variable because its value is shared across multiple functions. The.get value() and.set value() methods can be used to obtain and modify the value. In the method, there is an additional parameter called update. updates must be accompanied by a list of form pairs (shared-variable, new expression).

Complete Guide to Tensorflow for Deep Learning with Python for Free

For more related projects -

/projects/data-science-projects/tensorflow-projects
/projects/data-science-projects/keras-deep-learning-projects

Example -

# Import libraries
import theano
from theano import shared,tensor

# Creating shared variable as 'shared_variable' with value '0'
shared_variable = shared(0)

# Creating a scalar
x = tensor.iscalar('x')

# Creating function which will take x as input and returns shared_variable and later updates shared_variable value with some expression mentioned in updates parameter.
fun = theano.function([x], shared_variable, updates=[(shared_variable, shared_variable + x)])

# Calling function
print('Calling function =>',fun(10))

# Printing shared_variable value
print('Value of shared variable =>',shared_variable.get_value())

# Calling function
print('Calling function =>',fun(100))

# Printing shared_variable value
print('Value of shared variable =>',shared_variable.get_value())


Output -
Calling function => 0
Value of shared variable => 10
Calling function => 10
Value of shared variable => 110

In this way, we can use a shared variable in theano.