[R] bnlearn cpquery NA problem
yuj|@n| @end|ng |rom um|ch@edu
Wed Jun 12 19:34:16 CEST 2019
I'm using bnlearn to build a Bayesian Network with 167 nodes. I have got
the fitted network, but when I try to do inference using cpquery (or
cpdist), I got this error message:
Error in logical(n) : vector size cannot be NA/NaNIn addition: Warning
message:In check.cpq.args(fitted = fitted, event = NULL, extra.args =
list(...), : NaNs produced
I got the bn.fit object by calling bn.fit using "bayes" method, and I
also tried "mle" method with replace.unidentifiable setting to TRUE.
But both of them gave me the above error when doing inference.
Does anyone have any recommended solution to fix this?
[[alternative HTML version deleted]]
More information about the R-help