Skip to content

Commit 1b69ce4

Browse files
[Term Entry] PyTorch Tensor Operations: .logical_or()
* [Edit] Python: Python CLI arguments * Update command-line-arguments.md * [Term Entry] PyTorch Tensor Operations: .log2() * [Term Entry] PyTorch Tensor Operations: .logical_or() * Delete docs/content/pytorch/concepts/tensor-operations/terms/log2/log2.md * Rename docs/content/pytorch/concepts/tensor-operations/terms/logical-or/logical-or.md. to content/pytorch/concepts/tensor-operations/terms/logical-or/logical-or.md * Update logical-or.md ---------
1 parent 9701ac2 commit 1b69ce4

File tree

1 file changed

+100
-0
lines changed
  • content/pytorch/concepts/tensor-operations/terms/logical-or

1 file changed

+100
-0
lines changed
Lines changed: 100 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,100 @@
1+
---
2+
Title: '.logical_or()'
3+
Description: 'Computes the element-wise logical OR between two tensors.'
4+
Subjects:
5+
- 'Computer Science'
6+
- 'Data Science'
7+
Tags:
8+
- 'Booleans'
9+
- 'PyTorch'
10+
- 'Tensor'
11+
CatalogContent:
12+
- 'learn-python-3'
13+
- 'paths/data-science'
14+
---
15+
16+
The **`.logical_or()`** function performs an element-wise logical OR operation between two [tensors](https://www.codecademy.com/resources/docs/pytorch/tensors). Each input value is converted to a Boolean value (zero → `False`, non-zero → `True`) before the operation. The output is a Boolean tensor where each element is `True` if at least one of the corresponding input elements is `True`.
17+
18+
## Syntax
19+
20+
```pseudo
21+
torch.logical_or(input, other, out)
22+
```
23+
24+
**Parameters:**
25+
26+
- `input`: A tensor whose elements are treated as Boolean.
27+
- `other`: A tensor broadcastable with input.
28+
- `out` (optional): A tensor to store the result.
29+
30+
**Return value:**
31+
32+
Returns a Boolean tensor where each element is `True` if at least one of the corresponding input elements is `True`.
33+
34+
## Example 1
35+
36+
In this example, `.logical_or()` evaluates pairs of values from two 2-D tensors and returns a Boolean matrix showing which positions satisfy the OR condition:
37+
38+
```py
39+
import torch
40+
41+
a = torch.tensor([[0, 1], [2, 0]])
42+
b = torch.tensor([[3, 0], [0, 5]])
43+
44+
result = torch.logical_or(a, b)
45+
46+
print("Tensor A:\n", a)
47+
print("\nTensor B:\n", b)
48+
print("\nA OR B:\n", result)
49+
```
50+
51+
The output of this code is:
52+
53+
```shell
54+
Tensor A:
55+
tensor([[0, 1],
56+
[2, 0]])
57+
58+
Tensor B:
59+
tensor([[3, 0],
60+
[0, 5]])
61+
62+
A OR B:
63+
tensor([[True, True],
64+
[True, True]])
65+
```
66+
67+
## Example 2
68+
69+
In this example, a 1-D tensor is combined with a scalar using broadcasting to show how `.logical_or()` can apply a Boolean condition efficiently across all elements:
70+
71+
```py
72+
import torch
73+
74+
x = torch.tensor([0, 4, 0, 7])
75+
y = torch.tensor(1) # scalar
76+
77+
result = torch.logical_or(x, y)
78+
79+
print(result)
80+
```
81+
82+
The output of this code is:
83+
84+
```shell
85+
tensor([True, True, True, True])
86+
```
87+
88+
## Frequently Asked Questions
89+
90+
### 1. What are the tensor operations in PyTorch?
91+
92+
Tensor operations in PyTorch include arithmetic, logical operations, reductions, reshaping, indexing, broadcasting, and matrix operations. These operations run efficiently on CPU or GPU and form the core building blocks of neural network workflows.
93+
94+
### 2. Why use tensor instead of NumPy?
95+
96+
Tensors support automatic differentiation and execute seamlessly on GPUs, which makes them suited for deep learning workloads. They also integrate tightly with PyTorch's computation graph, while NumPy arrays operate only on the CPU and lack gradient support.
97+
98+
### 3. What does cpu() do in PyTorch?
99+
100+
The `.cpu()` method moves a tensor or model from a GPU device to the system’s CPU. This is useful for running operations on hardware without CUDA support or preparing data for libraries that operate only on CPU-based arrays.

0 commit comments

Comments
 (0)