how to apply a function to each group - r

Question

I have this function*,
x<- data.frame (
'Ones'=c(1,0,0,0,0,1,0,0,0,0,1,0,0,0,0,0,0,0,0,0),
'Thats'=c(0,5,3,6,8,4,5,6,8,3,1,NA,4,5,6,7,4,3,4,5))
f<- function(df, n){
df %>%
collapse::flag(-n:n) %>%
rowSums(na.rm = T) - 1
}
x %>%
mutate(gap1 = f(., 1),
gap2 = f(., 2),
gap3 = f(., 3),
gap4 = f(., 4),
gap5 = f(., 5)) %>%
filter(Ones == 1)
It finds when the variable Ones is qual to 1 and sums,
if gap1 = f(., 1), the values n-1, n and n+1 in Thats, with n being in the same row as 1,
if gap2 = f(.,2) n-2 + n-1 + n + n+1 + n+2...
And so on.
This is the output.
Ones Thats gap1 gap2 gap3 gap4 gap5
1 1 0 5 8 14 22 27
2 1 4 17 29 40 48 51
3 1 1 4 16 27 38 50
Now I want to use this function for each group of my sample.
This is a Facsimile of my sample. The real sample has more conditions and more observation, but this is just to give you an idea.
dat<- data.frame (
'sub' = c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2),
'trial' = c(1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2),
'Thumb'= c(1,0,0,0,0,1,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,1,0,0,0,0,0,0,1,0,0),
'Index'= c(0,5,3,6,8,4,5,6,8,3,1,3,4,5,6,7,4,3,4,5,4,5,3,23,2,1,4,6,3,2,3,2,4,6,3,2,4,3,2,1))
I need to apply the function to all groups of my sample.
For example, in this specific case I would 4 conditions:
sub1/trial1, sub2/trial1, sub1/trial2, sub2/trial2
I tried to write the function in this way:
f<- function(df, n){
df %>% group_by(trial, sub) %>%
collapse::flag(-n:n) %>%
rowSums(na.rm = T) - 1
}
But it doesn't work.
Do you have any advice to give me?
Please consider that the sample above is only an example... I need a general function working also with more conditions and observations.
Thank you,
Best regards
*credits: Kindly proposed by Donald Seinen on this site.

Answer 1

We may need group_modify if we are using the first function f
library(dplyr)
dat %>%
group_by(trial, sub) %>%
group_modify(~ .x %>%
mutate(gap1 = f(.x, 1), gap2 = f(.x, 2),
gap3 = f(.x, 3), gap4 = f(.x, 4), gap5 = f(.x, 5))) %>%
ungroup
-output
# A tibble: 40 × 9
trial sub Thumb Index gap1 gap2 gap3 gap4 gap5
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 1 1 0 5 8 14 22 27
2 1 1 0 5 8 14 22 27 32
3 1 1 0 3 13 22 27 32 38
4 1 1 0 6 16 26 32 38 46
5 1 1 0 8 18 26 37 46 49
6 1 1 1 4 17 29 40 48 49
7 1 1 0 5 15 31 40 43 48
8 1 1 0 6 18 26 34 40 43
9 1 1 0 8 16 21 26 34 40
10 1 1 0 3 10 16 21 26 34
# … with 30 more rows
which is the same result we get on the subset of data for the first combination
> dat %>%
filter(sub == 1, trial == 1) %>%
select(-sub, -trial) %>%
mutate(gap1 = f(., 1),
gap2 = f(., 2),
gap3 = f(., 3),
gap4 = f(., 4),
gap5 = f(., 5))
Thumb Index gap1 gap2 gap3 gap4 gap5
1 1 0 5 8 14 22 27
2 0 5 8 14 22 27 32
3 0 3 13 22 27 32 38
4 0 6 16 26 32 38 46
5 0 8 18 26 37 46 49
6 1 4 17 29 40 48 49
7 0 5 15 31 40 43 48
8 0 6 18 26 34 40 43
9 0 8 16 21 26 34 40
10 0 3 10 16 21 26 34