How to Filter Arrays in NumPy: Simple Guide with Examples

To filter a NumPy array, use boolean indexing by creating a condition inside array[condition]. This returns a new array with elements that meet the condition, like arr[arr > 5] to get values greater than 5.

Syntax

The basic syntax to filter a NumPy array is filtered_array = array[condition]. Here, condition is a boolean array of the same shape as array, where True means keep the element and False means discard it.

For example, array > 5 creates a boolean array marking elements greater than 5.

python

filtered_array = array[condition]

Example

This example shows how to filter elements greater than 5 from a NumPy array.

python

import numpy as np

arr = np.array([1, 3, 7, 9, 2, 5, 8])
filtered = arr[arr > 5]
print(filtered)

Output

[7 9 8]

Common Pitfalls

One common mistake is trying to filter using a condition without using boolean indexing, like arr > 5 alone, which only returns a boolean array, not the filtered values.

Another error is mixing data types or shapes, which causes errors when applying conditions.

python

import numpy as np

arr = np.array([1, 2, 3, 4])

# Wrong: This returns a boolean array, not filtered values
print(arr > 2)

# Right: Use boolean indexing to get filtered values
print(arr[arr > 2])

Output

[False False  True  True]
[3 4]

Quick Reference

• Boolean indexing: Use array[condition] to filter.
• Conditions: Use comparison operators like >, <, ==, !=.
• Combine conditions: Use & (and), | (or) with parentheses, e.g. arr[(arr > 2) & (arr < 8)].

Key Takeaways

Use boolean indexing with conditions to filter NumPy arrays.

Conditions must produce boolean arrays matching the original array shape.

Combine multiple conditions with & and | inside parentheses.

Filtering returns a new array with only elements meeting the condition.

Avoid using conditions alone without indexing, as they return booleans, not filtered data.