I have a data frame named “dat” with 5 numeric variables (var1, var2,var3,var4 , var5), each with 20 observations.
structure(list(var_1 = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 16, 17, 18, 19, 20), var_2 = c(7, 8, 9, 10, 11, 12,
13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26), var_3 = c(4,
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21,
22, 23), var_4 = c(2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
15, 16, 17, 18, 19, 20, 21)), row.names = c(NA, -20L), class = c("tbl_df",
"tbl", "data.frame"))
I am using this code to create and save into a new data frame (named “combined”) all possible pairs of combination of the 5 variables with the mean value of the 2 variables which are combined together:
combined <- combn(dat, 2, FUN = rowMeans)
This is the result:
structure(c(4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17,
18, 19, 20, 21, 22, 23, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 8.5, 9.5,
10.5, 11.5, 12.5, 13.5, 14.5, 15.5, 16.5, 17.5, 18.5, 19.5, 20.5,
21.5, 1.5, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 8.5, 9.5, 10.5, 11.5,
12.5, 13.5, 14.5, 15.5, 16.5, 17.5, 18.5, 19.5, 20.5, 5.5, 6.5,
7.5, 8.5, 9.5, 10.5, 11.5, 12.5, 13.5, 14.5, 15.5, 16.5, 17.5,
18.5, 19.5, 20.5, 21.5, 22.5, 23.5, 24.5, 4.5, 5.5, 6.5, 7.5,
8.5, 9.5, 10.5, 11.5, 12.5, 13.5, 14.5, 15.5, 16.5, 17.5, 18.5,
19.5, 20.5, 21.5, 22.5, 23.5, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 16, 17, 18, 19, 20, 21, 22), .Dim = c(20L, 6L))
1) The code works fine, but the problem is that the new generated variables in the data frame “combined”, are named V1,V2,V3,V4…. and I cannot understand each new variable which combination of variables comes from. I would prefer that the new generated variables will be named as “var1var2”, “var1var3” and so on… Is there a way to obtain this?
2) Also, is there a way to apply the combn function only to some columns and no to all variables which are present in the dataframe “dat”?
3) How can I add the new generated variables in the original data frame “dat” rather saving them in a new one?
Thank you so much for your help!
This can be done by redoing the combn statement with the column names
set.seed(99)
dat <- data.frame(var1 = sample(20), #some sample data
var2 = sample(20), #I did this before you added your data above!
var3 = sample(20),
var4 = sample(20),
var5 = sample(20))
dat
var1 var2 var3 var4 var5
1 12 5 18 19 12
2 3 2 10 6 13
3 13 15 14 13 1
4 17 11 16 18 10
5 9 13 8 8 7
6 15 6 20 17 3
...
combined <- combn(dat, 2, FUN = rowMeans) #your statement using cols of dat
colnames(combined) <- combn(names(dat), 2, paste0, collapse="") #same using colnames
combined
var1var2 var1var3 var1var4 var1var5 var2var3 var2var4 var2var5 var3var4 var3var5 var4var5
[1,] 8.5 15.0 15.5 12.0 11.5 12.0 8.5 18.5 15.0 15.5
[2,] 2.5 6.5 4.5 8.0 6.0 4.0 7.5 8.0 11.5 9.5
[3,] 14.0 13.5 13.0 7.0 14.5 14.0 8.0 13.5 7.5 7.0
[4,] 14.0 16.5 17.5 13.5 13.5 14.5 10.5 17.0 13.0 14.0
[5,] 11.0 8.5 8.5 8.0 10.5 10.5 10.0 8.0 7.5 7.5
[6,] 10.5 17.5 16.0 9.0 13.0 11.5 4.5 18.5 11.5 10.0
...
To answer your other points, you can restrict the columns used by using, for example dat[,c(2,3,6)] in the comb statements (to use columns 2, 3 and 6). You can add them back to the same dataframe with dat <- cbind(dat, combined)