How to Use np.argwhere in NumPy: Syntax and Examples

Use np.argwhere to find the indices of elements in a NumPy array that meet a condition. It returns a 2D array where each row is the index of a matching element. This is useful for locating positions of elements that satisfy a boolean expression.

Syntax

The basic syntax of np.argwhere is:

• np.argwhere(condition)

Here, condition is a boolean array or expression applied to a NumPy array. The function returns a 2D array of indices where the condition is True.

python

np.argwhere(a > 0)

Example

This example shows how to use np.argwhere to find indices of elements greater than 3 in a 2D array.

python

import numpy as np

a = np.array([[1, 4, 2], [3, 5, 0]])
indices = np.argwhere(a > 3)
print(indices)

Output

[[0 1]
 [1 1]]

Common Pitfalls

One common mistake is expecting np.argwhere to return a flat list of indices for 1D arrays. Instead, it always returns a 2D array of indices, where each row is a coordinate. For 1D arrays, you may want to flatten the result using .flatten() or use np.where for simpler output.

Also, passing a non-boolean array or incorrect condition will cause errors or unexpected results.

python

import numpy as np

arr = np.array([10, 20, 30, 40])

# Wrong: expecting flat indices directly
wrong_indices = np.argwhere(arr > 15)
print(wrong_indices)  # Output is 2D

# Right: flatten to get 1D indices
right_indices = np.argwhere(arr > 15).flatten()
print(right_indices)

# Alternative: use np.where for 1D arrays
where_indices = np.where(arr > 15)[0]
print(where_indices)

Output

[[1]
 [2]
 [3]]
[1 2 3]
[1 2 3]

Quick Reference

Function	Description
np.argwhere(condition)	Returns 2D array of indices where condition is True
np.argwhere(condition).flatten()	Flattens indices to 1D array (useful for 1D arrays)
np.where(condition)	Returns tuple of arrays of indices, often simpler for 1D arrays

Key Takeaways

np.argwhere returns a 2D array of indices where the condition is True.

For 1D arrays, flatten the result or use np.where for simpler index lists.

Always pass a boolean condition to np.argwhere to avoid errors.

np.argwhere is useful for locating positions of elements matching any condition in multi-dimensional arrays.