How to perform element-wise subtraction on tensors in PyTorch?

In this article, we are going to understand how to perform element-wise subtraction on tensors in PyTorch in Python. We can perform element-wise subtraction using torch.sub() method.

torch.sub() method allows us to perform subtraction on the same or different dimensions of tensors. It takes two tensors as the inputs and returns a new tensor with the result (element-wise subtraction). If tensors are different in dimensions so it will return the higher dimension tensor. we can also subtract a scalar quantity with a tensor using torch.sub() function. We can use the below syntax to compute the element-wise subtraction.

Syntax: torch.sub(input, other, *, alpha=1, out=None)

Parameters:

• input: the input tensor.
• other: This is tensor or number to subtract from the input tensor.
• alpha (Number): the parameter is multiplier for other.
• out: it is the output tensor, This is optional parameter.

Return: it will returns a new modified tensor with element-wise subtraction of the tensor input by the tensor other.

Example 1:

The following program is to perform element-wise subtraction on two single dimension tensors.

Python3

# import torch library
import torch

# define two tensors
tens_1 = torch.Tensor([10, 20, 30, 40, 50])
tens_2 = torch.Tensor([1, 2, 3, 4, 5])

# display tensors
print(" First Tensor: ", tens_1)
print(" Second Tensor: ", tens_2)

# multiply tensors
tens = torch.sub(tens_1, tens_2)

# display result after perform element wise
# subtraction
print(" After Element-wise subtraction: ", tens)

Output:

Example 2:

The following program is to perform element-wise subtraction on two 2D tensors.

Python3

# import torch library
import torch

# define two tensors
tens_1 = torch.Tensor([[10, 20], [30, 40]])

tens_2 = torch.Tensor([[1, 2], [3, 4]])

# display tensors
print("First Tensor:", tens_1)
print("Second Tensor:", tens_2)

# Subtract tensors
tens = torch.sub(tens_1, tens_2)

# display result after perform element wise
# subtraction
print("After Element-wise subtraction:", tens)

Output:

Example 3:

The following program is to know how to subtract a scalar quantity to a tensor.

Python3

# import torch library
import torch

# define a tensor
tens = torch.Tensor([100, 200, 300, 400, 500])

# display tensors
print(tens)

# Subtract 50 from tensor
tens_result = torch.sub(tens, 50)

# display result after subtract scalar from tensor
print("After subtract scalar from tensor:", tens_result)

Output:

Example 4:

The following program is to understand how to perform element-wise subtraction on two different dimension tensors.

Python3

# import torch library
import torch

# define 2D tensor
tens_1 = torch.Tensor([[100, 200], [300, 400]])

# define 1D tensor
tens_2 = torch.Tensor([10, 20])

# display tensors
print("First Tensor:", tens_1)
print("Second Tensor:", tens_2)

# Subtract tensors
tens = torch.sub(tens_1, tens_2)

# display result after perform element wise subtraction
print("After Element-wise subtraction:", tens)

Output: