Errors when using bridgesampling

I fit a gaussian process model and want to evaluate marginal likelihood through bridgesampling package. However, it returns error:

>   marg.list[[1]] <- bridge_sampler(fit, m, method = "warp3", 
+                                    repetitions=1, cores=1, verbose=TRUE)

[1] "summary(q12): (log_dens of proposal for posterior samples)"
      Min.    1st Qu.     Median       Mean    3rd Qu.       Max. 
-2.709e+09 -2.225e+09 -2.144e+09 -2.147e+09 -2.058e+09 -1.135e+09 
[1] "summary(q22): (log_dens of proposal for generated samples)"
[[1]]
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  -3062   -2984   -2963   -2963   -2941   -2851 

[1] "summary(q11): (log_dens of posterior for posterior samples)"
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
 -76756  -76667  -76645  -76645  -76621  -76555 
[1] "summary(q21): (log_dens of posterior for generated samples)"
[[1]]
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
 -82434  -76886  -76786  -76869  -76712  -76522 

Iteration: 1
Iteration: 2
Iteration: 1
Error in jj[2, ] : subscript out of bounds
In addition: Warning messages:
1: Infinite value in iterative scheme, returning NA.
 Try rerunning with more samples. 
2: logml could not be estimated within maxiter, rerunning with adjusted starting value. 
Estimate might be more variable than usual. 

Do you know what causes the problem and whether there are ways to solve it? I could provide fit and data if needed (however, it’s a bit large). Thank you!

Edit: included backticks for the code (by @Max_Mantei)

Hey there! I don’t know whats happening here. But @Henrik_Singmann is on here sometimes, maybe he knows what’s happening or who to ask for help?

Cheers,
Max

1 Like

Hi Max, thanks for your response. I have contacted Henrik and it’s fine now. Thank you!

1 Like

That’s great to hear! Do you mind sharing the solution here so others might benefit from it as well?

1 Like

Hi, actually the author didn’t come across this problem before, but he does suggest to increase number of iterations due to the large parameter space.

1 Like