For loop with random variables different person

Does anyone have an idea create "FOR" loop in R that would repeat all variables choice n times and the only thing that changes is human id (after 16384 rows, where human id = 1, plus another 16384 rows should be with humanid = 2). I tried to create new data frame where a$humanid<-rep(2,nrow(a)) and combine two tables, but that would be stupid, since I need 100 humans with same variables and different choice.

Here is my code for variables with humanid=1:

a<-expand.grid(price75 = seq(1,0), price100 = seq(1,0),
               price125 = seq(1,0), fuel5 = seq(1,0), fuel15 = seq(1,0),
               fuel25 = seq(1,0), co2_50 = seq(1,0), co2_75 = seq(1,0),
               co2_100 = seq(1,0), range400 = seq(1,0), range700 = seq(1,0),
               range1000 = seq(1,0), avail60 = seq(1,0), avail100 = seq(1,0))
a$humanid<-rep(1,nrow(a)) 

a$choice<-rep(0,nrow(a))
a$choice <- `[<-`(integer(length(a$choice)), sample(length(a$choice), 1), 1)
a$alt <- 1:nrow(a)

Thanks in advance.

CodePudding user response：

If I understand the question correctly, a for loop is not necessary. Add humanid to the expand.grid, then take a row sample for each ID, then offset those in order to get each in the correct set of 2^14 rows.

a<-expand.grid(price75 = seq(1,0), price100 = seq(1,0),
               price125 = seq(1,0), fuel5 = seq(1,0), fuel15 = seq(1,0),
               fuel25 = seq(1,0), co2_50 = seq(1,0), co2_75 = seq(1,0),
               co2_100 = seq(1,0), range400 = seq(1,0), range700 = seq(1,0),
               range1000 = seq(1,0), avail60 = seq(1,0), avail100 = seq(1,0),
               humanid = 1:100, choice = 0)
a$choice[sample(2^14, 100, TRUE)   (0:99)*2^14] <- 1

CodePudding user response：

Uhm.. Assume there are m "real" rows, and you want to replicate up to n ones (so assume m <= n; Then you can do modulo arithmetic:

m <- 10
n <- 25

(seq_len(n) - 1) %% 10   1
 [1]  1  2  3  4  5  6  7  8  9 10  1  2  3  4
[15]  5  6  7  8  9 10  1  2  3  4  5

Where you need the -1 and 1 because of R indexing from 1 instead of 0.