python-pytorchHow can I use Python and PyTorch to reshape a tensor?

The reshape function in the PyTorch library is used to reshape a tensor. It takes a tuple as an argument, which specifies the desired shape of the tensor. The following example shows how to use reshape to reshape a tensor from a 3x3 matrix to a 9x1 vector.

import torch

# Create a 3x3 matrix
x = torch.tensor([[1, 2, 3],
                  [4, 5, 6],
                  [7, 8, 9]])

# Reshape to a 9x1 vector
y = x.reshape(9, 1)

print(y)

Output example

tensor([[1],
        [2],
        [3],
        [4],
        [5],
        [6],
        [7],
        [8],
        [9]])

Code explanation

• import torch: imports the PyTorch library.
• x = torch.tensor([[1, 2, 3], [4, 5, 6], [7, 8, 9]]): creates a 3x3 tensor.
• y = x.reshape(9, 1): reshapes the tensor to a 9x1 vector.
• print(y): prints the reshaped tensor.

Helpful links

• PyTorch Documentation - Tensor Reshaping

Edit this code on GitHub