Aggregation Functions in NumPy
sum(), min(), max(), mean(), and More

Introduction to Aggregation in NumPy

Aggregation functions are the backbone of data analysis in NumPy. Whether you're summing up monthly sales, finding the maximum temperature from a dataset, or calculating the mean of sensor readings — aggregation functions provide the answers quickly and efficiently.

What Are Aggregation Functions?

Aggregation functions reduce an entire array to a single value or perform summary statistics along a specified axis. NumPy offers a wide variety of such functions, including:

• np.sum() – Total of all elements
• np.min() and np.max() – Minimum and maximum values
• np.mean() – Average value
• np.std() and np.var() – Standard deviation and variance
• np.argmin() and np.argmax() – Index of the minimum/maximum value

Creating a Sample Array

Let’s begin with a basic 2D NumPy array to explore these functions:

import numpy as np

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

[[4 7 1]
 [8 5 9]]

Using np.sum() – Add Everything

This function adds up all the values in the array.

print(np.sum(arr))

34

The sum of all values: 4 + 7 + 1 + 8 + 5 + 9 = 34

Sum by Axis

print(np.sum(arr, axis=0))  # Column-wise sum
print(np.sum(arr, axis=1))  # Row-wise sum

[12 12 10]
[12 22]

Explanation: For axis=0, we are summing columns: [4+8, 7+5, 1+9]. For axis=1, we’re summing rows: [4+7+1], [8+5+9].

Minimum and Maximum Values

To find the smallest and largest values:

print(np.min(arr))
print(np.max(arr))

1
9

Min/Max Along Axis

print(np.min(arr, axis=1))  # Min of each row
print(np.max(arr, axis=0))  # Max of each column

[1 5]
[8 7 9]

Mean – The Average

To find the mean of all elements:

print(np.mean(arr))

5.666666666666667

The average of 6 numbers in the array: 34 / 6 = 5.666...

Standard Deviation and Variance

These help understand the spread of your data. Standard deviation shows how far values are from the mean.

print(np.std(arr))
print(np.var(arr))

2.6246692913372702
6.888888888888889

Finding Index of Extremes: argmin and argmax

print(np.argmin(arr))  # Index in flattened array
print(np.argmax(arr))

2
5

Index 2 corresponds to the value 1, and index 5 is 9 in the flattened array.

Things to Watch Out For

• If an array contains np.nan, aggregation results may also be nan. Use np.nansum(), np.nanmean(), etc. to handle such cases.
• When working with multidimensional arrays, always be explicit about the axis parameter to avoid confusion.
• Operations return new values — original arrays remain unchanged unless explicitly reassigned.

Recap

Aggregation functions in NumPy give you quick, clear insights into your data. Whether you need a simple total or a detailed statistical spread, these tools are essential for any kind of numerical computing. Practice each function, play with the axes, and understand the shape and behavior of the results to become truly confident with array operations.