Displaying 2 results from an estimated 2 matches for "howmanymiss".
2011 Jun 01
1
Missing completely at random
...lt;- 12 ####number of variables
X <- matrix(rnorm(N * n), N, n) ####matrix
pMiss <- 0.05 ####percent of missing values
idMiss <- sample(1:N, N * pMiss) ####sample cases
nMiss <- length(idMiss)
m <- 3 ####maximum number of missing variables within selected cases
howmanyMiss <- sapply(idMiss, function(x) sample(1:m, 1))
howmanyMiss
lapply(howmanyMiss, function(x) sample(1:n, x))
Any suggestion?
Thanks,
Blaž
[[alternative HTML version deleted]]
2011 Jun 03
3
Not missing at random
Hello!
I would like to sample 30 % of cases (with at least 1 value lower than 3) and
among them I want to set all values lower than 3 (within selected cases) as NA
(NMAR- Not missing at random). I managed to sample cases, but I don’t know how
to set values (lower than 3) as NA.
R code:
x <-