np.sum() and axis parameter in NumPy

Introduction

We use np.sum() to add up numbers in arrays quickly. The axis parameter helps us choose how to add: across rows, columns, or the whole array.

Adding all sales numbers in a table to get total sales.

Summing scores of each student across subjects to get total marks.

Calculating total pixels brightness in each row or column of an image.

Finding total expenses per category from a budget spreadsheet.

Syntax

NumPy

np.sum(array, axis=None)

array is the data you want to add up.

axis=None sums all elements into one number.

axis=0 sums down columns (vertical).

axis=1 sums across rows (horizontal).

Examples

Adds every number in the array to get one total.

NumPy

np.sum(arr)
# sums all elements in arr

Adds numbers down each column, returns sums per column.

NumPy

np.sum(arr, axis=0)
# sums each column

Adds numbers across each row, returns sums per row.

NumPy

np.sum(arr, axis=1)
# sums each row

Sample Program

This code creates a 3x3 array. It sums all numbers, then sums each column, and then sums each row. It prints all results clearly.

NumPy

import numpy as np

arr = np.array([[1, 2, 3],
                [4, 5, 6],
                [7, 8, 9]])

# Sum all elements
total_sum = np.sum(arr)

# Sum down columns (axis=0)
sum_columns = np.sum(arr, axis=0)

# Sum across rows (axis=1)
sum_rows = np.sum(arr, axis=1)

print(f"Total sum: {total_sum}")
print(f"Sum of each column: {sum_columns}")
print(f"Sum of each row: {sum_rows}")

Important Notes

If you forget axis, it adds everything into one number.

For 2D arrays, axis=0 means down columns, axis=1 means across rows.

For higher dimensions, axis chooses which dimension to sum over.

Summary

np.sum() adds numbers in arrays.

The axis parameter controls the direction of addition.

Use axis=0 for columns, axis=1 for rows, or None for all elements.

Practice

1. What does the axis parameter in np.sum() control?

easy

A. The data type of the output array.

B. The direction along which the sum is calculated (rows or columns).

C. The size of the array to sum.

D. The number of times the sum is repeated.

Solution

1. Step 1: Understand the role of axis in np.sum()

   The axis parameter tells numpy whether to sum across rows (axis=1) or columns (axis=0).

2. Step 2: Clarify what axis does not affect

   It does not change data type, size, or repetition of sums.

3. Final Answer:

   The direction along which the sum is calculated (rows or columns). -> Option B

4. Quick Check:

   axis controls sum direction [OK]

Hint: Axis decides row or column sum direction [OK]

Common Mistakes:

• Thinking axis changes data type
• Confusing axis with array size
• Assuming axis repeats sums

2. Which of the following is the correct syntax to sum all elements of a 2D numpy array arr?

easy

A. np.sum(arr)

B. np.sum(arr, axis=1, 0)

C. np.sum(arr, axis=2)

D. np.sum(arr, axis='all')

Solution

1. Step 1: Recall default behavior of np.sum()

   Without axis, np.sum(arr) sums all elements in the array.

2. Step 2: Check invalid axis values

   axis=2 is invalid for 2D arrays, axis=1,0 is incorrect syntax, axis='all' is invalid.

3. Final Answer:

   np.sum(arr) -> Option A

4. Quick Check:

   Default np.sum sums all elements [OK]

Hint: No axis sums all elements [OK]

Common Mistakes:

• Using invalid axis values
• Passing multiple axis values incorrectly
• Using string instead of integer for axis

3. Given the array arr = np.array([[1, 2, 3], [4, 5, 6]]), what is the output of np.sum(arr, axis=0)?

medium

A. [5 7 9]

B. [3 7]

C. [6 15]

D. [1 2 3 4 5 6]

Solution

1. Step 1: Understand axis=0 sums columns

   Summing along axis=0 adds elements vertically: (1+4), (2+5), (3+6).

2. Step 2: Calculate sums for each column

   1+4=5, 2+5=7, 3+6=9, so result is [5 7 9].

3. Final Answer:

   [5 7 9] -> Option A

4. Quick Check:

   Sum columns = [5 7 9] [OK]

Hint: axis=0 sums columns vertically [OK]

Common Mistakes:

• Confusing axis=0 with axis=1
• Adding rows instead of columns
• Expecting a scalar output

4. What is wrong with this code snippet?

import numpy as np
arr = np.array([[1, 2], [3, 4]])
result = np.sum(arr, axis=2)

medium

A. The array is not defined correctly.

B. np.sum() cannot be used on 2D arrays.

C. The axis value 2 is invalid for a 2D array.

D. The import statement is missing.

Solution

1. Step 1: Check array dimensions

   The array has shape (2, 2), so valid axis values are 0 or 1.

2. Step 2: Validate axis parameter

   axis=2 is out of bounds for this 2D array, causing an error.

3. Final Answer:

   The axis value 2 is invalid for a 2D array. -> Option C

4. Quick Check:

   Axis must be less than array dimensions [OK]

Hint: Axis must be less than array dimensions [OK]

Common Mistakes:

• Using axis larger than array dimensions
• Thinking np.sum can't handle 2D arrays
• Ignoring error messages

5. You have a 3D numpy array arr = np.array([[[1,2],[3,4]], [[5,6],[7,8]]]). What is the shape of the result when you run np.sum(arr, axis=1)?

hard

A. (1, 2)

B. (3, 2)

C. (2, 4)

D. (2, 2)

Solution

1. Step 1: Understand the shape of arr

   arr has shape (2, 2, 2): 2 blocks, each with 2 rows and 2 columns.

2. Step 2: Summing along axis=1

   axis=1 sums over the second dimension (rows), reducing it from 2 to 1, so shape changes from (2, 2, 2) to (2, 2).

3. Final Answer:

   (2, 2) -> Option D

4. Quick Check:

   Sum axis=1 reduces dimension 2 to 1 [OK]

Hint: Sum along axis removes that dimension [OK]

Common Mistakes:

• Confusing which axis is summed
• Expecting shape to stay same
• Mixing up dimensions in 3D arrays