bas.lm and bas.glm ignoring prior model probabilities that are 0

[merliseclyde]

Describe the bug
BAS is including models in the output that have prior probability 0.

To Reproduce
Steps to reproduce the behavior:

data("bodyfat")
bas_mod <- bas.lm(Bodyfat ~.,data = bodyfat[1:14,], method = 'BAS', modelprior = tr.poisson(2, 5))
pred <- predict(bas_mod,newdata = bodyfat[15:20,], se.fit = T, estimator = 'BMA') 
cbind(bas_mod$rank, bas_mod$size, bas_mod$postprobs)

Expected behavior

Expect that model with zero prior and posterior probability are dropped from the output and calculations

Desktop (please complete the following information):

• OS: OSX
• R Version 4.3.2
• BAS 1.7.0

Additional context
related to Issue #70

[merliseclyde]

partial fix of issue using postprocessing of output to drop all models with prior probability zero. A better solution is to eliminate the models from the tree-structure as they are sampled in C with methods BAS, deterministic and MCMC+BAS as is done with MCMC.