How to Find Unique Values in NumPy Arrays Easily

Use the np.unique() function to find unique values in a NumPy array. It returns sorted unique elements and can also return counts or indices if needed.

Syntax

The basic syntax of np.unique() is:

• np.unique(array, return_index=False, return_inverse=False, return_counts=False, axis=None)

Here:

• array: Input NumPy array to find unique values from.
• return_index: If True, returns indices of first occurrences of unique values.
• return_inverse: If True, returns indices to reconstruct original array from unique values.
• return_counts: If True, returns counts of each unique value.
• axis: If specified, finds unique rows or columns along this axis.

python

np.unique(array, return_index=False, return_inverse=False, return_counts=False, axis=None)

Example

This example shows how to find unique values in a 1D NumPy array and get their counts.

python

import numpy as np

arr = np.array([1, 2, 2, 3, 4, 4, 4, 5])
unique_values, counts = np.unique(arr, return_counts=True)
print("Unique values:", unique_values)
print("Counts:", counts)

Output

Unique values: [1 2 3 4 5]
Counts: [1 2 1 3 1]

Common Pitfalls

Some common mistakes when using np.unique() include:

• Forgetting that the returned unique values are sorted, which may change the original order.
• Not using return_counts=True when counts are needed.
• Trying to find unique rows in a 2D array without specifying axis=0.

Example of wrong and right usage:

python

import numpy as np

arr_2d = np.array([[1, 2], [1, 2], [3, 4]])

# Wrong: unique without axis returns unique elements flattened
print(np.unique(arr_2d))

# Right: unique rows by specifying axis=0
print(np.unique(arr_2d, axis=0))

Output

[1 2 3 4]
[[1 2]
 [3 4]]

Quick Reference

Summary of np.unique() options:

Parameter	Description
array	Input array to find unique values from
return_index	Return indices of first unique values (default False)
return_inverse	Return indices to reconstruct original array (default False)
return_counts	Return counts of each unique value (default False)
axis	Find unique rows/columns along this axis (default None)

Key Takeaways

Use np.unique() to get sorted unique values from a NumPy array.

Set return_counts=True to get how many times each unique value appears.

Specify axis=0 or axis=1 to find unique rows or columns in 2D arrays.

Remember np.unique() returns sorted results, which may reorder data.

Use return_index or return_inverse to map unique values back to original array.