Randomly increase values of a row (binary), with specific conditions - r

Question

I have a data set like this:
A B C D E F G
12 1 0 0 0 0 0
Hey 0 1 0 0 0 0
No 0 0 0 0 0 1
Yes 0 0 0 0 1 0
I want to build an scenario, what will happen if a COLUMN has 10% more YES (yes = 1). But, in my scenario, this should be done with 3 columns at the same time.
So: let's say that the rows of interest are where B=1 or C=1 or D=1. If one of the columns is equal to 1, that is fine. But I want to (randomly) make 10% of the remaining rows (where B=0 & C=0 & D=0) into a 1 (and of course, if we give them (the randomly 10% of the remaining rows) a 1, then the other columns should be all 0 (except column A)).
Sorry, really had a hard time to explain this problem. Hopefully it is clear.
The result should be something like this (it is not representing the 10% since the example is too small).
A B C D E F G
12 1 0 0 0 0 0
Hey 0 1 0 0 0 0
No 0 0 0 0 0 1
Yes 0 1 0 0 0 0
where you can see that "Yes" is randomly assigned as C=1, and its original value is set back to 0.

Answer 1

I believe this is what you want:
data:
df1<-
structure(list(A = c("12", "Hey", "No", "Yes"), B = c(1L, 0L,
0L, 0L), C = c(0L, 1L, 0L, 0L), D = c(0L, 0L, 0L, 0L), E = c(0L,
0L, 0L, 0L), F = c(0L, 0L, 0L, 1L), G = c(0L, 0L, 1L, 0L)), row.names = c(NA,
-4L), class = "data.frame")
code:
m <- `rownames<-`(df1[,-1],df1[,1]) # make your life simple, add character col as rownames
percentage = .5 # choose any percentage you like from 0 to 1, .1 for 10%
amountOf1 = floor(percentage * ncol(m)) # get the amount of ones based on percentage
IND <- which(rowSums(m[,1:3]) == 0) # get those rows having B, C, D with 0
for(i in IND) {
m[i,] = sample(rep(1:0,c(amountOf1,ncol(m)-amountOf1)) )
}
result: (now 50% are 1 in rows where B,C,D is 0)
# B C D E F G
#12 1 0 0 0 0 0
#Hey 0 1 0 0 0 0
#No 1 0 0 0 1 1
#Yes 1 0 1 0 0 1