How to Compute row means

R Function
Learn how to compute row means with this comprehensive R tutorial. Includes practical examples and code snippets.
Published

February 24, 2023

Introduction

Computing row means allows you to calculate the average value across columns for each row in your dataset. This is particularly useful when you have multiple measurements per observation and want to create summary statistics or composite scores.

Getting Started

library(tidyverse)
library(palmerpenguins)

Example 1: Basic Usage

The Problem

We need to calculate the average of numeric columns across each row in a data frame. Let’s start with a simple dataset to understand the fundamental approach.

Step 1: Create sample data

We’ll create a small dataset with test scores to demonstrate row means calculation.

# Create sample test scores
test_scores <- data.frame(
  student = c("Alice", "Bob", "Carol"),
  math = c(85, 92, 78),
  science = c(88, 87, 82),
  english = c(91, 89, 85)
)

This creates a dataset where each row represents a student and each column (except student) contains their test scores.

Step 2: Calculate row means using base R

The rowMeans() function provides the simplest way to calculate row averages.

# Calculate row means for numeric columns
test_scores$average <- rowMeans(test_scores[, c("math", "science", "english")])
print(test_scores)

The rowMeans() function automatically computes the mean across specified columns for each row, giving us each student’s average score.

Step 3: Handle missing values

When your data contains NA values, you need to specify how to handle them.

# Create data with missing values
test_scores_na <- test_scores
test_scores_na$math[2] <- NA

# Calculate means ignoring NA values
test_scores_na$average <- rowMeans(test_scores_na[, 2:4], na.rm = TRUE)
print(test_scores_na)

The na.rm = TRUE parameter ensures missing values are excluded from the calculation rather than resulting in NA.

Example 2: Practical Application

The Problem

Let’s work with the Palmer penguins dataset to calculate average body measurements for each penguin. This represents a real-world scenario where you might want to create a composite measure from multiple related variables.

Step 1: Prepare the penguin data

We’ll select relevant measurement columns and remove any rows with missing values.

# Prepare penguin measurement data
penguin_measurements <- penguins |>
  select(species, bill_length_mm, bill_depth_mm, 
         flipper_length_mm, body_mass_g) |>
  filter(complete.cases(.))

This gives us a clean dataset with four measurement variables for each penguin.

Step 2: Standardize measurements before averaging

Since our measurements are on different scales, we should standardize them before computing means.

# Standardize measurements (z-scores)
penguin_std <- penguin_measurements |>
  mutate(across(bill_length_mm:body_mass_g, 
                ~ scale(.)[,1], 
                .names = "std_{.col}"))

Standardization converts each measurement to z-scores, making them comparable across different units and scales.

Step 3: Calculate row means using dplyr

We’ll use dplyr’s rowwise() and c_across() functions for a modern approach.

# Calculate average standardized measurement
penguin_avg <- penguin_std |>
  rowwise() |>
  mutate(avg_measurement = mean(c_across(starts_with("std_")))) |>
  ungroup()

This approach provides flexibility to select columns using helper functions like starts_with() or contains().

Step 4: Compare species averages

Let’s see how our row means vary across penguin species.

# Summarize by species
species_summary <- penguin_avg |>
  group_by(species) |>
  summarise(
    mean_composite = mean(avg_measurement),
    sd_composite = sd(avg_measurement),
    .groups = "drop"
  )
print(species_summary)

This reveals how the composite measurement differs between Adelie, Chinstrap, and Gentoo penguins.

Summary

  • Use rowMeans() for simple row mean calculations across numeric columns
  • Always specify na.rm = TRUE when dealing with missing values in your data
  • Consider standardizing variables before computing means when measurements use different scales
  • Use rowwise() with c_across() in dplyr for more flexible column selection