Tensorflow.js tf.divNoNan() Function

Tensorflow.js is an open-source library developed by Google for running machine learning models and deep learning neural networks in the browser or node environment.

The tf.divNoNan() function is used to divide two Tensors element-wise and returns 0 if the denominator is 0. It supports broadcasting.

Syntax:

tf.divNoNan (a, b)

Parameters: This function accepts two parameters which are illustrated below:

• a: The first input tensor as the numerator.
• b: The second input tensor as the denominator. It should have the same data type as “a”.

Return Value: It returns a Tensor for the result of a/b, where a is the first Tensor and b is the second Tensor. It returns 0 if the denominator is 0.

Example 1:

Javascript

// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"

// Initializing some Tensors
const a = tf.tensor1d([2, 5, 7, 10]);
const b = tf.tensor1d([1, 3, 2, 6]);
const c = tf.tensor1d([0, 0, 0, 0]);

// Calling the .divNoNan() function
// over the above Tensors as its parameters
a.divNoNan(b).print();
a.divNoNan(c).print();

Output:

Tensor
   [2, 1.6666665, 3.5, 1.6666665]
Tensor
   [0, 0, 0, 0]

Example 2:

Javascript

// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"

// Broadcasting div a with b and c
const a = tf.tensor1d([3, 6, 11, 17]);
const b = tf.scalar(2);
const c = tf.scalar(0);

// Calling the .divNoNan() function
// over the above Tensors as its parameters
a.divNoNan(b).print();
a.divNoNan(c).print();

Output:

Tensor
   [1.5, 3, 5.5, 8.5]
Tensor
   [0, 0, 0, 0]