bugs in getConditionallyIndependentSets when `givenNodes` provided

[paciorek]

  mc <- nimbleCode({
    mu ~ dnorm(0,1)
    for(i in 1:2) {
      x[i] ~ dnorm(mu, 1)
      y[i] ~ dnorm(x[i], 1)
      z[i] ~ dnorm(y[i], 1)
    }
  })
  m <- nimbleModel(mc, data = list(z = 1:2))
m$getConditionallyIndependentSets(givenNodes = 'z')

[[1]]
[1] "x[1]" "y[1]"

[[2]]
[1] "x[2]" "y[2]"

but if one doesn't condition on mu then they are not independent.

Also:

m$getConditionallyIndependentSets(givenNodes = 'mu')

Error: SEXP_2_vectorInt called for SEXP that is not a numeric or logica!
Error: We could not handle input type to SEXP_2_vectorInt

[paciorek]

Perry pointed out that because of how unknownAsGiven is used and what nodes is by default, the first is known behavior and not a bug.