Filter grouped data with filter function and if else condition in R

Question

Consider the data frame

id <-c(1,1,1,2,2,3,3,3,3)
x1 <-c("no","yes","yes","no","no","no","no","no","yes")
x2 <-c(9,3,7,4,5,2,5,6,11)
df <- data.frame(id, x1, x2)

For each id I wish to keep the first row where x1="yes" and where x1 is completely "no" for a particular id I want to retain the last row for that id. I wrote the code

df %>% group_by(id) %>% 
       filter(if(x1 == "yes") x1 == "yes" else row_number() %in% c(n()))

but it runs into an error. The expect is

id  x1 x2
 1 yes  3
 2  no  5
 3 yes 11

I was wondering if this code could be improved to get this issue resolved? I also appreciate alternative approaches (e.g., data.table approach) of handling this. Thank you!

Simon · Accepted Answer · 2025-07-29 06:29:39Z

4

your code fails because filter produces a vector for each group instead of a single true/false.
so x1 == "yes" produces a vector like c(FALSE, TRUE, TRUE) while your if statement expects a singular TRUE or FALSE.

I recommend using slice for this use case as such:

df %>%
  group_by(id) %>%
  slice(if (any(x1 == "yes")) which.max(x1 == "yes") else n())

any(x1 == "yes") makes it so that it checks the entire group and returns a single TRUE or FALSE
and using slice() picks rows by their position (eg: 1st row, 5th row)

so which.max(x1 == "yes") finds the position of the first "yes" and selects it.
and if none exist, it gives the position of the last "no"

answered Jul 29 at 6:29

Simon

4187 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

s_baldur · Accepted Answer · 2025-07-30 08:40:17Z

3

Using data.table:

library(data.table)
setDT(df)
df[, .SD[match("yes", x1, nomatch = .N)], by = id]

#       id     x1    x2
#    <num> <char> <num>
# 1:     1    yes     3
# 2:     2     no     5
# 3:     3    yes    11

Same in dplyr:

slice(df, match("yes", x1, nomatch = n()), .by = id)

Your original solution with minor modifications

df %>% group_by(id) %>% filter(row_number() == min(n(), which(x1 == "yes")))

Make sure assumptions hold

stopifnot(all(df$x1 %in% c("yes", "no")))

edited Jul 30 at 8:40

answered Jul 29 at 12:35

s_baldur

34.5k4 gold badges43 silver badges80 bronze badges

Collectives™ on Stack Overflow

Filter grouped data with filter function and if else condition in R

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related