PyTorch .reciprocal()
Anonymous contributor
Published Oct 30, 2025
Contribute to Docs
The .reciprocal() method performs an element-wise operation that returns the multiplicative inverse of every element in the tensor. For any element x, the result is 1/x. This method works with tensors that have floating-point or complex data types (such as torch.float32, torch.float64, or torch.complex64). If an element equals zero, the reciprocal for that position becomes infinity (inf).
Syntax
torch.reciprocal(input, *, out=None)
Parameters:
input(Tensor): The input tensor whose elements will be inverted.out(Tensor, optional): The output tensor to store the result. Must have the same shape asinput.
Return value:
Returns a tensor containing the element-wise reciprocals of input.
Example 1: Compute element-wise reciprocals
In this example, we create a tensor of floating-point numbers and calculate their reciprocals using torch.reciprocal():
import torch# Create a tensor of floating-point valuesx = torch.tensor([2.0, 4.0, 0.5, 0.0])print("Original Tensor:")print(x)# Compute the reciprocalresult = torch.reciprocal(x)print("\nReciprocal Tensor:")print(result)# Note: Division by zero results in 'inf'
The output of this code is:
Original Tensor:tensor([2.0000, 4.0000, 0.5000, 0.0000])Reciprocal Tensor:tensor([0.5000, 0.2500, 2.0000, inf])
Example 2: Using the out parameter
In this example, the result is stored in a pre-allocated output tensor using the out parameter:
import torch# Input tensorx = torch.tensor([1.0, 2.0, 0.25])# Create an empty tensor with the same shapeout_tensor = torch.empty_like(x)# Store the result in 'out_tensor'torch.reciprocal(x, out=out_tensor)print("Input Tensor:")print(x)print("\nReciprocal stored in 'out' tensor:")print(out_tensor)
The output of this code is:
Input Tensor:tensor([1.0000, 2.0000, 0.2500])Reciprocal stored in 'out' tensor:tensor([1.0000, 0.5000, 4.0000])
All contributors
- Anonymous contributor
Contribute to Docs
- Learn more about how to get involved.
- Edit this page on GitHub to fix an error or make an improvement.
- Submit feedback to let us know how we can improve Docs.
Learn PyTorch on Codecademy
- Machine Learning Data Scientists solve problems at scale, make predictions, find patterns, and more! They use Python, SQL, and algorithms.
- Includes 27 Courses
- With Professional Certification
- Beginner Friendly.95 hours
- Learn how to use PyTorch to build, train, and test artificial neural networks in this course.
- Intermediate.3 hours