#0037

Normalize Each Row

MediumNumPy18 min15 XP

Problem

Given a 2D NumPy array, divide each row by that row sum. If a row sum is zero, leave that row as all zeros.

Why This Matters

This pattern appears in probabilities, attention weights, transition matrices, and feature normalization.

Function Signature

def normalize_rows(matrix):

Examples

Example 1

Inputmatrix = np.array([[1, 1], [2, 6], [0, 0]])

Outputarray([[0.5, 0.5], [0.25, 0.75], [0.0, 0.0]])

Each nonzero row is divided by its own row sum. The zero row stays zero.

Constraints

Input is a 2D numeric NumPy array.
Use row-wise sums, not one global sum.
Avoid division by zero for all-zero rows.

CodePython

import numpy as np

def normalize_rows(matrix):
    """
    Input example: matrix = np.array([[1, 1], [2, 6], [0, 0]])
    Return: array([[0.5, 0.5], [0.25, 0.75], [0.0, 0.0]])
    Return the value from the function so the tests can compare it. Do not print unless the prompt asks.
    """
    # write your code here
    return matrix

import numpy as np

def normalize_rows(matrix):
    """
    Input example: matrix = np.array([[1, 1], [2, 6], [0, 0]])
    Return: array([[0.5, 0.5], [0.25, 0.75], [0.0, 0.0]])
    Return the value from the function so the tests can compare it. Do not print unless the prompt asks.
    """
    # write your code here
    return matrix

Visible browser tests run here when available.

Testcases1 visible / 4 hidden categories

Nonzero and zero rows

Inputnp.array([[1, 1], [2, 6], [0, 0]])

Expected[[0.5, 0.5], [0.25, 0.75], [0.0, 0.0]]

keepdims=True keeps row sums shaped for broadcasting.

Hidden Test Categories

One-row matrixRows with decimal valuesRows with negative and positive valuesMany columns