I have dataframe with 6 columns and I want to partition them into 3 parts each with 2 columns. I want to recombine partitions in all possible combinations to create 7 new dataframes
part1,part2,part3
part1,par2
part1,part3
part2,part3
part1
part2
part3
I modified this solution a bit to recombine them Split a dataframe into all possible combinations of dataframes by 3 columns in R
>frame <- data.frame(id = letters[seq( from = 1, to = 10 )], a = rnorm(10, 4), b = rnorm(10, 6), c=rnorm(10, 5), d = rnorm(10, 2),e=rnorm(10, 5), f = rnorm(10, 2))
> frame
id a b c d e f
1 a 6.322845 5.828619 5.465636 2.7658092 6.522706 1.4896078
2 b 2.352437 5.521230 6.555715 0.6612871 5.288508 2.4837969
3 c 2.790967 9.253197 3.724231 2.9954273 4.887744 1.3020424
4 d 2.017975 6.038846 4.540511 1.7989492 6.059974 -0.2463154
5 e 4.004463 4.384898 5.341084 1.9528288 4.186449 1.0823939
6 f 2.600336 6.562758 5.708489 2.1142707 6.769220 1.7942291
7 g 3.850400 7.231973 4.918542 3.3562489 6.090841 1.4202527
8 h 2.932744 6.377516 5.518261 1.7423230 4.422915 1.8789437
9 i 5.135185 5.218992 4.710196 1.1878825 5.421876 0.8455756
10 j 5.188278 7.233590 6.303500 0.3868047 4.390973 1.6997801
>m <- seq(3)
>j <-function(m){lapply(as.data.frame(combn(ncol(frame) - 1, m)), function(idx) frame[, c(1, idx + 1)])}
>lapply(m, function(m) j(m))
This would create all combinations by shuffling all columns. I do not want combinations of columns, but combinations of partitions. How can I achieve that?
Here is one try:
library(dplyr)
library(purrr)
# Assign a partition to be used here
# (Updated from OP's clarification about pttns & @bouncyball's comment)
pttn <- split(names(frame)[-1], rep(1:3, each = 2))
# Create combinations of partitioned columns
do.call(c, lapply(seq_along(pttn), combn, x = pttn, simplify = FALSE)) %>%
map(~ frame %>% select(reduce(.x, c)))
The first line with do.call creates all combinations of 'partitions', or the partitioned column names. If you want to preserve ID column, you can use id, reduce(.x, c) instead of reduce(.x, c)