dplyr Mutate

Introduction

The mutate() function creates new columns based on existing ones. It's essential for feature engineering.

Basic Mutate

library(dplyr)

df <- tibble(
  x = 1:5,
  y = c(10, 20, 30, 40, 50)
)

# Create new column
mutate(df, sum = x + y)

# Multiple new columns
mutate(df, 
       sum = x + y,
       product = x * y,
       ratio = y / x)

Using in mutate()

df <- tibble(
  name = c("Alice", "Bob", "Charlie"),
  score1 = c(85, 90, 78),
  score2 = c(80, 95, 88)
)

mutate(df,
       average = (score1 + score2) / 2,
       difference = score1 - score2,
       status = ifelse(average >= 85, "Pass", "Fail"))

Mutate with Window Functions

df <- tibble(
  value = c(10, 20, 30, 40, 50)
)

mutate(df,
       lag_value = lag(value, 1),
       lead_value = lead(value, 1),
       cumulative = cumsum(value),
       rank = rank(value))

Transmute

# Transmute keeps only new columns
transmute(df,
          sum = x + y,
          product = x * y)

Summary

mutate() creates new variables from existing data. Use window functions for complex transformations.