numpy.sum() in Python

numpy.sum() is a NumPy function used to calculate the sum of array elements. It can sum values across the entire array or along a specific axis. It also allows controlling the output data type, initial value and shape of the result.

30

Explanation: np.sum(arr) adds all elements (5 + 10 + 15) and returns the total.

Syntax:

numpy.sum(arr, axis=None, dtype=None, out=None, initial=0, keepdims=False)

Parameters:

• arr: Input array whose elements are to be summed.
• axis: Axis along which the sum is computed. 0 -> column-wise, 1 -> row-wise and None -> entire array
• dtype: Data type of the returned sum.
• out: Output array to store the result.
• initial: Starting value of the sum.
• keepdims: Keeps the reduced axis as dimension in result.

Example 1: This example shows how numpy.sum() works on a 1D array and how changing the dtype affects the result.

36.2
36
36.2

Explanation:

• np.sum(arr) computes the sum using NumPy’s default data type, preserving decimal values.
• np.sum(arr, dtype=np.uint8) forces the result into an 8-bit unsigned integer, which supports values only from 0 to 255.
• Decimal values are discarded, and sums exceeding this range overflow and wrap around modulo 256.
• np.sum(arr, dtype=np.float32) performs the sum using floating-point arithmetic, preserving precision.

Note: Using small integer data types such as uint8 may produce unexpected results due to overflow, not calculation errors.

Example 2: This example calculates the sum of a 2D array and shows how using different data types changes the output.

279
23
279.0

Example 3: This example demonstrates summing a 2D array along rows, columns, and using keepdims=True.

279
[52 25 93 42 67]
[120 75 84]
[[120]
 [ 75]
 [ 84]]

Explanation:

• np.sum(arr, axis=0) column-wise sum.
• np.sum(arr, axis=1) row-wise sum.
• np.sum(arr, axis=1, keepdims=True) preserves the reduced dimension, returning a column-shaped result.