np.clip(arr, lo, hi) leaves elements already between lo and hi untouched, replaces any element below lo with lo, and replaces any element above hi with hi — a fast, vectorized way to enforce a valid range on an array's values without writing an explicit loop or boolean-indexing assignment. Passing None for either bound leaves that side unconstrained, letting you clip only a lower bound, only an upper bound, or both.

1Understanding np.clip()

💡

Use np.clip() to enforce a valid value range in one call, such as keeping pixel values within 0-255 or probabilities within 0-1, instead of writing two separate boolean-indexing assignments for the lower and upper bound.

editor.html

import numpy as np

arr = np.array([-5, 0, 5, 10, 15])
print(np.clip(arr, 0, 10))

localhost:3000

2Practical Example

Here is a real-world application of np.clip() showing how it is used in production NumPy code.

editor.html

import numpy as np

pixels = np.array([-20, 100, 260, 128])
print(np.clip(pixels, 0, 255))

localhost:3000

3Best Practices

Follow these guidelines when working with np.clip():

1. Use np.clip() instead of two separate boolean-indexing assignments when constraining values to a range

2. Pass None for one bound when you only need to enforce a floor or a ceiling, not both

3. Clip intermediate results in numerically unstable calculations, like probabilities that should stay strictly between 0 and 1, to avoid downstream errors from values that drift slightly out of a valid range

⚠️

Tip: Use np.clip() to enforce a valid value range in one call, such as keeping pixel values within 0-255 or probabilities within 0-1, instead of writing two separate boolean-indexing assignments for the lower and upper bound.

editor.html

import numpy as np

arr = np.array([-5, 0, 5, 10, 15])
print(np.clip(arr, 0, 10))

localhost:3000

Examples

Example 01Basic Usage

import numpy as np

arr = np.array([-5, 0, 5, 10, 15])
print(np.clip(arr, 0, 10))

Example 02Advanced Example

import numpy as np

pixels = np.array([-20, 100, 260, 128])
print(np.clip(pixels, 0, 255))

Best Practices

Use np.clip() instead of two separate boolean-indexing assignments when constraining values to a range
Pass None for one bound when you only need to enforce a floor or a ceiling, not both
Clip intermediate results in numerically unstable calculations, like probabilities that should stay strictly between 0 and 1, to avoid downstream errors from values that drift slightly out of a valid range

Interview Question

How is np.clip(arr, 0, 10) different from just filtering out values outside 0-10 with boolean indexing?▼

Hint: Think about what happens to the array's length and shape in each case.

np.clip() preserves the array's original shape and length exactly, replacing out-of-range values with the nearest boundary value rather than removing them — every original position still has a value in the result. Boolean indexing to filter out values outside a range instead removes those elements entirely, typically returning a shorter, flattened 1D array with a different length than the original. The two operations solve genuinely different problems: clip() constrains values in place, while filtering excludes elements from the result altogether.

Exercises

MediumPractice using np.clip() in a real scenario.

View Solution

import numpy as np

arr = np.array([-5, 0, 5, 10, 15])
print(np.clip(arr, 0, 10))

Frequently Asked Questions

How is np.clip(arr, 0, 10) different from just filtering out values outside 0-10 with boolean indexing?▼

Related Functions

boolean-indexingnp-minnp-max