How to Use round in NumPy: Syntax and Examples

Use numpy.round(array, decimals) to round elements of a NumPy array to a specified number of decimal places. The decimals parameter controls how many digits to keep after the decimal point, defaulting to 0 for rounding to the nearest integer.

Syntax

The numpy.round function rounds each element in an array to the specified number of decimals.

• array: The input NumPy array or number to round.
• decimals: Number of decimal places to round to (default is 0).

python

numpy.round(array, decimals=0)

Example

This example shows how to round a NumPy array of floats to 2 decimal places and to the nearest integer.

python

import numpy as np

arr = np.array([3.14159, 2.71828, 1.61803])
rounded_two_decimals = np.round(arr, 2)
rounded_integer = np.round(arr)

print("Original array:", arr)
print("Rounded to 2 decimals:", rounded_two_decimals)
print("Rounded to nearest integer:", rounded_integer)

Output

Original array: [3.14159 2.71828 1.61803]
Rounded to 2 decimals: [3.14 2.72 1.62]
Rounded to nearest integer: [3. 3. 2.]

Common Pitfalls

One common mistake is forgetting that decimals defaults to 0, so np.round(arr) rounds to integers. Another is expecting np.round to modify the original array; it returns a new array instead. Also, rounding floating-point numbers can sometimes produce unexpected results due to binary representation.

python

import numpy as np

arr = np.array([2.675, 2.685])

# Wrong: expecting in-place change
np.round(arr, 2)
print("Array after np.round call:", arr)

# Right: assign the result
rounded = np.round(arr, 2)
print("Rounded array:", rounded)

Output

Array after np.round call: [2.675 2.685]
Rounded array: [2.67 2.69]

Quick Reference

Remember these tips when using numpy.round:

• Input: array or number
• decimals: number of digits after decimal (can be negative to round to tens, hundreds, etc.)
• Returns: new array with rounded values
• Does not modify original array

Key Takeaways

Use numpy.round(array, decimals) to round array elements to desired decimal places.

The decimals parameter defaults to 0, rounding to nearest integer.

numpy.round returns a new array; it does not change the original array.

Negative decimals round to tens, hundreds, etc., e.g., decimals=-1 rounds to nearest 10.

Floating-point rounding can have small precision quirks due to binary representation.