Nested if and max if statement in R error: unused argument

Question

Still quite new to R and I would like to identify the biggest value of a variable within a group when an other variable is postive/negative. Specifically, if l_diff<0, I would like to find the biggest value of t in each group when pos_n<0. If l_diff>0, I would like to find the biggest value of t in each group when pos_p>0. Example data:

l_diff <- c(-1,-1,-1,-1,-1,-1,1,1,1,1,1)
pos_n <- c(2,2,1,-4,-2,-2,2,1,-5,4,8)
pos_p <- c(3,4,-5,6,-7,2,-3,3,2,1,4)
t <- c(5,7,3,1,6,2,7,5,3,2,1)
group <- c(1,1,1,1,1,1,2,2,2,2,2)
db <- data.frame(cbind(l_diff,pos_n, pos_p, t, group))

Desired output:

cmax<- c(6,6,6,6,6,6,5,5,5,5,5)

I tried the following:

db<-db %>%
  group_by((group)) %>%
  mutate(ifelse(l_diff<0, t1 = max(t[pos_n<0], ifelse(l_diff>0, t1 = max(t[pos_p >0])))))

But I get the following error:

Error: Problem with mutate() input ..1. x unused argument (t1 = max(t[pos_n < 0], ifelse(l_diff > 0, t1 = max(t[pos_p > 0])))) i Input ..1 is ifelse(...). i The error occurred in group 1: (group) = 1.

Any idea what may be wrong or any other suggestions?

Answer 1

With ifelse, we need to place the assignment outside, similarly, all the arguments in the ifelse usage must be present

ifelse(test, yes, no)

Here, the no was not found in the nested second ifelse. It is not an issue if we use case_when as by default the TRUE ~ NA

library(dplyr)
db %>%
   group_by(group) %>%
   mutate(t1  = ifelse(l_diff<0,  max(t[pos_n<0]), 
                ifelse(l_diff>0,max(t[pos_p >0]), NA))) %>%
   ungroup

-output

# A tibble: 11 x 6
#   l_diff pos_n pos_p     t group    t1
#    <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
# 1     -1     2     3     5     1     6
# 2     -1     2     4     7     1     6
# 3     -1     1    -5     3     1     6
# 4     -1    -4     6     1     1     6
# 5     -1    -2    -7     6     1     6
# 6     -1    -2     2     2     1     6
# 7      1     2    -3     7     2     5
# 8      1     1     3     5     2     5
# 9      1    -5     2     3     2     5
#10      1     4     1     2     2     5
#11      1     8     4     1     2     5