Question

I have a matrix of one column and 6 rows. I would like to replicate that column i times but change one value randomly each time, and after each iteration, calculate the mean and variance across all columns.

For example:

values = rnorm(6, 6, 1); matrix1 = matrix(values, 6)

After i=1, would look like:

values2 = values
values2[sample(1:6, 1)] = values2[sample(1:6, 1)]+runif(1, 0, 1)
matrix2 = matrix(c(values, values2), 6)

At the end, I would like to output a data frame that looks like so:

i   mean   var
1   1.23  2.31
2   1.24  2.33 etc...

For many i‘s. I imagine there is a way to do this with loops, but my skills are not such that I can figure it out. Thanks for all your help!

Answer 1

If you know how many times you’re doing this, it would be best to construct your final matrix beforehand, especially if i is large. However, without that:

jitter.func <- function(x, vec) {
  cell <- sample(1:length(vec), 1)
  vec[cell] <- vec[cell] + runif(1, 0, 1)

  return(c(mean=mean(vec), var=var(vec)))
}


i <- 10
sapply(1:i, jitter.func, vec=values)