Problem with mutate_all function in dplyr-CodePudding

I am not sure if this is a silly question, but I need a hint now. Apologies in advance.

So I have a random dataframe df and I want to use the mutate_at() function.

x1<-rpois(20,5)
x2<-rpois(20,5)
x3<-rpois(20,2)
df1<-data.frame(x1,x2,x3)

This work perfectly fine if I divide the whole dataframe by one column like this:

library(dplyr)
df1%>%mutate_at(vars(x1:x3),list(All_by_x3=~./x3))

But what I actually want, is that x3 is divided by each column of the dataframe. So I want x3 as the numerator and loop through the different columns in the denominator. What I logically tried is this:

df1%>%mutate_at(vars(x1:x3),list(All_by_x3= x3/ ~.))

but it doesn't work. It feels like there is a simple answer, but I can't see it.

Thanks for any help.

CodePudding user response：

A possible solution, based on dplyr:

library(dplyr)

df1 %>% 
  mutate(across(x1:x2, ~ x3/.x, .names = "x3_{.col}"))

#>    x1 x2 x3     x3_x1      x3_x2
#> 1   2  4  1 0.5000000 0.25000000
#> 2   7  3  2 0.2857143 0.66666667
#> 3   8 11  1 0.1250000 0.09090909
#> 4   7  4  2 0.2857143 0.50000000
#> 5   3  7  0 0.0000000 0.00000000
#> 6   6  6  3 0.5000000 0.50000000
#> 7   2  7  2 1.0000000 0.28571429
#> 8   7  6  1 0.1428571 0.16666667
#> 9   1  5  0 0.0000000 0.00000000
#> 10  4  7  6 1.5000000 0.85714286
#> 11  3  4  2 0.6666667 0.50000000
#> 12  4  3  2 0.5000000 0.66666667
#> 13  2  1  2 1.0000000 2.00000000
#> 14  2  4  0 0.0000000 0.00000000
#> 15  6  4  1 0.1666667 0.25000000
#> 16  5  5  4 0.8000000 0.80000000
#> 17  7  5  6 0.8571429 1.20000000
#> 18  5  4  3 0.6000000 0.75000000
#> 19  6  5  3 0.5000000 0.60000000
#> 20  5  3  1 0.2000000 0.33333333