This function uses MCMC samples from fitted htlrfit
object OR user supplied
regression coefficient to predict the class labels of test cases.
htlr_predict( X_ts, fithtlr = NULL, deltas = NULL, burn = NULL, thin = 1, usedmc = NULL, rep.legacy = TRUE )
X_ts | Matrix of values at which predictions are to be made. |
---|---|
fithtlr | Fitted HTLR model object. |
deltas | The values of deltas (for example true deltas) used to make prediction;
will override |
burn, thin |
|
usedmc | Indices of Markov chain iterations used for inference.
If supplied, |
rep.legacy | To reproduce (actually incorrect) results in legacy version. See https://github.com/longhaiSK/HTLR/issues/7. |
A matrix of predictive probabilities, with rows for cases, cols for classes.