← SASnRconditional logic - if_else - nested conditions

Introduction

When a derivation needs three or more categories — say, under 20, 20-59, and 60+ — we can nest if/else calls instead of reaching for case_when.
SAS stacks if/else if/else clauses one after another in a DATA step.
Tidyverse nests if_else() inside the no slot of an outer if_else(). Base R does the same with ifelse().
For more than 3-4 branches, case_when() (tidyverse) becomes noticeably easier to read.
This lesson bands age into three categories and assigns the label to agegr1.
After running, rows fall into ">= 60 Years", "20 - < 60 Years", or "< 20 Years".

conditional logic - if_else - nested conditions

R (tidyverse)+

library(dplyr)

# Sample data
dm <- tibble(
  USUBJID = c("101-001", "101-002", "101-003", "101-004"),
  age = c(35, 64, 19, 17)
)

dm02 <- dm %>%
  mutate(agegr1 = if_else(age >= 60, ">= 60 Years",
    if_else(between(age, 20, 59), "20 - < 60 Years", "< 20 Years")
  ))

For three-branch logic in tidyverse, we can nest if_else() calls — the "else" value of the outer call is itself another if_else().
The outer if_else(age >= 60, ..., ...) handles the 60+ case. The inner if_else(between(age, 20, 59), ..., ...) handles the 20-59 and everything-else cases.
between(age, 20, 59) is a dplyr helper equivalent to age >= 20 & age <= 59 — a small readability boost for the simple case.
Each branch must return the same type, same as the non-nested if_else().
For more branches, prefer case_when() — see the case_when lesson (1865). Nesting gets hard to read past three levels.

Happy learning!

Search lessons

Introduction

conditional logic - if_else - nested conditions