np.argmin() and np.argmax() in NumPy

Introduction

These functions help find the position of the smallest or largest value in data. This is useful to quickly locate important points in your data.

Finding the day with the lowest temperature in a week.

Identifying the highest score in a test among students.

Locating the minimum or maximum sales value in a monthly report.

Determining the index of the fastest runner in a race.

Finding the position of the cheapest product in a price list.

Syntax

NumPy

np.argmin(array, axis=None)
np.argmax(array, axis=None)

array is the data you want to check.

axis lets you choose if you want the position in the whole array or along rows/columns.

Examples

Finds the index of the smallest value in a simple list.

NumPy

np.argmin([3, 1, 2])

Finds the index of the largest value in a simple list.

NumPy

np.argmax([3, 1, 2])

Finds the index of the smallest value in each column.

NumPy

np.argmin([[5, 2, 3], [1, 4, 6]], axis=0)

Finds the index of the largest value in each row.

NumPy

np.argmax([[5, 2, 3], [1, 4, 6]], axis=1)

Sample Program

This program uses np.argmin() and np.argmax() to find positions of lowest and highest scores in a 2D array of student test scores. It shows how to find these positions overall and along rows or columns.

NumPy

import numpy as np

# Create a 2D array representing scores of 3 students in 4 tests
scores = np.array([[88, 92, 79, 93],
                   [85, 90, 91, 89],
                   [90, 85, 88, 92]])

# Find the index of the lowest score overall
lowest_score_index = np.argmin(scores)

# Find the index of the highest score overall
highest_score_index = np.argmax(scores)

# Find the index of the lowest score in each test (column-wise)
lowest_per_test = np.argmin(scores, axis=0)

# Find the index of the highest score for each student (row-wise)
highest_per_student = np.argmax(scores, axis=1)

print(f"Lowest score index overall: {lowest_score_index}")
print(f"Highest score index overall: {highest_score_index}")
print(f"Lowest score index per test (column-wise): {lowest_per_test}")
print(f"Highest score index per student (row-wise): {highest_per_student}")

Important Notes

Indexes start at 0, so the first element is at position 0.

If you do not specify axis, the function treats the array as flat (1D).

These functions return the first occurrence if there are multiple equal min or max values.

Summary

np.argmin() finds the position of the smallest value.

np.argmax() finds the position of the largest value.

You can find positions overall or along rows/columns using the axis parameter.

Practice

1. What does the function np.argmin() do in NumPy?

easy

A. Finds the index of the smallest value in an array

B. Finds the index of the largest value in an array

C. Returns the smallest value in an array

D. Returns the largest value in an array

Solution

1. Step 1: Understand the purpose of np.argmin()

   The function np.argmin() returns the index (position) of the smallest value in a NumPy array.

2. Step 2: Differentiate from similar functions

   Unlike np.min() which returns the smallest value itself, np.argmin() returns where that value is located.

3. Final Answer:

   Finds the index of the smallest value in an array -> Option A

4. Quick Check:

   np.argmin() = index of smallest value [OK]

Hint: Remember: argmin = position of minimum value [OK]

Common Mistakes:

• Confusing np.argmin() with np.min()
• Thinking it returns the smallest value, not the index
• Mixing up np.argmin() with np.argmax()

2. Which of the following is the correct NumPy function syntax to find the index of the largest value in a NumPy array arr?

easy

A. np.argmin(arr)

B. arr.argmax()

C. np.argmax(arr)

D. arr.argmin()

Solution

1. Step 1: Recall the function to find largest value index

   np.argmax(arr) returns the index of the largest value in the array arr.

2. Step 2: Check other options

   arr.argmax() is valid but not the function syntax asked; np.argmin() and arr.argmin() find smallest value index, not largest.

3. Final Answer:

   np.argmax(arr) -> Option C

4. Quick Check:

   np.argmax(arr) = index of max value [OK]

Hint: Use np.argmax() to get largest value index [OK]

Common Mistakes:

• Using np.argmin() instead of np.argmax()
• Confusing method call arr.argmax() with function np.argmax()
• Mixing up largest and smallest value functions

3. Given the array arr = np.array([3, 1, 4, 1, 5]), what is the output of np.argmin(arr)?

medium

A. 3

B. 1

C. 0

D. 4

Solution

1. Step 1: Identify the smallest value in the array

   The array is [3, 1, 4, 1, 5]. The smallest value is 1.

2. Step 2: Find the index of the first occurrence of the smallest value

   The first 1 appears at index 1 (0-based indexing).

3. Final Answer:

   1 -> Option B

4. Quick Check:

   np.argmin([3,1,4,1,5]) = 1 [OK]

Hint: np.argmin() returns first smallest value index [OK]

Common Mistakes:

• Choosing the last occurrence index instead of first
• Confusing values with indices
• Using np.argmax() by mistake

4. Consider the code:

import numpy as np
arr = np.array([[2, 5, 1], [7, 3, 9]])
index = np.argmax(arr, axis=1)

What is the correct output of index?

medium

A. [1 2]

B. [0 2]

C. [2 1]

D. [2 2]

Solution

1. Step 1: Understand np.argmax with axis=1

   axis=1 means find the index of the largest value in each row.

2. Step 2: Find largest values per row and their indices

   Row 0: [2, 5, 1], max is 5 at index 1; Row 1: [7, 3, 9], max is 9 at index 2.

3. Step 3: Check options carefully

   [2 2] shows [2 2], which is incorrect because first row max index is 1, not 2. [1 2] is [1 2], which matches the correct indices.

4. Final Answer:

   [1 2] -> Option A

5. Quick Check:

   np.argmax(arr, axis=1) = [1 2] [OK]

Hint: axis=1 finds max index per row [OK]

Common Mistakes:

• Mixing up axis parameter
• Confusing row and column indices
• Assuming global max index instead of per row

5. You have a 2D NumPy array representing exam scores of 4 students in 3 subjects:

scores = np.array([[88, 92, 79], [95, 85, 91], [70, 78, 88], [90, 91, 85]])

How do you find the subject index where each student scored the lowest?

hard

A. np.argmax(scores, axis=0)

B. np.argmin(scores, axis=0)

C. np.argmax(scores, axis=1)

D. np.argmin(scores, axis=1)

Solution

1. Step 1: Understand the array and axis meaning

   Each row is a student, columns are subjects. axis=1 means operate across columns (subjects) per student.

2. Step 2: Use np.argmin with axis=1 to find lowest score subject per student

   np.argmin(scores, axis=1) returns the index of the smallest score in each student's row.

3. Final Answer:

   np.argmin(scores, axis=1) -> Option D

4. Quick Check:

   Lowest subject per student = np.argmin(scores, axis=1) [OK]

Hint: Use axis=1 to find min per row (student) [OK]

Common Mistakes:

• Using axis=0 which finds min per subject, not per student
• Using np.argmax instead of np.argmin
• Confusing rows and columns in axis parameter