How to Calculate Row-Wise Statistics in R
Calculate row-wise statistics when each observation needs a summary value computed from its own columns — a student’s average across exams, a product total across sales channels, or a row maximum for outlier detection. R’s vectorized functions work down columns by default, so you need to explicitly tell R to work across columns instead. For performance, base R’s rowMeans() and rowSums() are the fastest option — they are vectorized C functions that outperform any R-level loop. Use apply(df, 1, f) only when the statistic you need is not covered by one of the built-in row functions.
df <- data.frame(
math = c(85, 92, 78),
science = c(90, 88, 82),
english = c(88, 85, 79)
)
df$avg <- rowMeans(df)
df$total <- rowSums(df)
df$max <- apply(df, 1, max)rowMeans() and rowSums() are significantly faster than apply(df, 1, mean) on large data frames. For more complex aggregations, dplyr::rowwise() pairs with c_across():
library(dplyr)
df %>%
rowwise() %>%
mutate(avg = mean(c_across(everything())))Both approaches handle NAs, but you must opt in. rowMeans() and rowSums() accept na.rm = TRUE, and rowwise() passes na.rm through to each summary function:
df$avg <- rowMeans(df[, c("math", "science")], na.rm = TRUE)If you do not set na.rm = TRUE and any column has NA, the result for that row is also NA. Use complete.cases(df) first if you want to exclude rows with any missing value before computing statistics.