Tensorflow.js tf.logSumExp() Function
Last Updated :
18 May, 2021
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.logSumExp() function is used to calculate log sum exp of elements of a Tensor across its dimension. It reduces the given input elements along the dimensions of axes. If the parameter “keepDims” is true, the reduced dimensions are retained with length 1 else the rank of Tensor is reduced by 1. If the axes parameter has no entries, it returns a Tensor with a single element with all reduced dimensions.
Syntax:
tf.logSumExp (x, axis, keepDims)
Parameters: This function accepts three parameters which are illustrated below:
- x: The input tensor.
- axis: The specified dimension(s) to reduce. By default it reduces all dimensions. It is optional parameter.
- keepDims: If this parameter value is true, it retains reduced dimensions with length 1 else the rank of Tensor is reduced by 1. It is also optional parameter.
Return Value: It returns a Tensor of calculated value of log sum exp operation.
Example 1:
Javascript
import * as tf from "@tensorflow/tfjs"
const a = tf.tensor1d([0, 1]);
const b = tf.tensor1d([3, 5]);
const c = tf.tensor1d([2, 4, 7]);
a.logSumExp().print();
b.logSumExp().print();
c.logSumExp().print();
|
Output:
Tensor
1.31326162815094
Tensor
5.126927852630615
Tensor
7.054985046386719
Example 2:
Javascript
import * as tf from "@tensorflow/tfjs"
const a = tf.tensor1d([0, 1]);
const b = tf.tensor2d([3, 5, 2, 8], [2, 2]);
const c = tf.tensor1d([2, 4, 7]);
const axis1 = -1;
const axis2 = -2;
const axis3 = 0;
a.logSumExp(axis1).print();
b.logSumExp(axis2, true ).print();
c.logSumExp(axis1, false ).print();
b.logSumExp(axis3, false ).print();
|
Output:
Tensor
1.31326162815094
Tensor
[[3.3132617, 8.0485878],]
Tensor
7.054985046386719
Tensor
[3.3132617, 8.0485878]
Reference:https://js.tensorflow.org/api/latest/#logSumExp
Like Article
Suggest improvement
Share your thoughts in the comments
Please Login to comment...