CODEWITHSUNDEEP
rsapply
user9292
user9292

Reputation: 1145

Error with sapply in R

I'm generating 100 random numbers from Dirichlet distribution, and then I need to use the output to generate from Gamma distribution. Here is the code:

a <- rdirichlet(100, c(1,1,1))
b <- c(3,3,3)

sapply(a, function(x) {rgamma(100, shape=2, rate =(b%*%a)) })

Note here that the rate for gamma distribution is the dot product of vectors b and a (which is the output from Dirichlet).

I'm getting this error message:

Error in b %*% a : non-conformable arguments

Upvotes: 0

Views: 697

Answers (1)

IRTFM
IRTFM

Reputation: 263301

I suspect you want. My library has at least four different packages with rdirichlet functions.):

library(MCMCpack)
apply(a,  1, function(x) {rgamma(100, shape=2, rate =(b %*% x)) })

When vectors are passed to %*% then need to be the same length and sapply was passing single elements rather than length-3 rows. (You also didn't have "x" in the expression.)

c(3,3,3) %*% 1
#Error in c(3, 3, 3) %*% 1 : non-conformable arguments

 str(a %*% b) 
 # vectors can be assumed to be column matrices in the second position
 num [1:100, 1] 3 3 3 3 3 3 3 3 3 3 ...

OR: 

> str(b %*% t(a))
# .... or assumed to be row matrices in the first position.
 num [1, 1:100] 3 3 3 3 3 3 3 3 3 3 ...

Upvotes: 1

Related Questions