search

dmphillippo / multinma

Network meta-analysis of individual and aggregate data in Stan
https://dmphillippo.github.io/multinma
33 stars 15 forks source link

Apparent error using binomial_2par likelihood model and loo with moment matching #30

Open MilaniC opened 8 months ago

MilaniC commented 8 months ago

The following error arises when invoking:

loo::loo(albie_fit_RE2, moment_match = TRUE)

to address some high Pareto k diagnostic values where albie_fit_RE2 is otherwise a success nma fit object.

_Error: Exception: variable does not exist; processing stage=parameter initialization; variable name=beta_tilde; base type=double (in 'binomial2par', line 202, column 0 to column 22)

[UPDATE: This error occurs irrespective of likelihood chosen]

This works:

loo::loo(albie_fit_RE2, moment_match = FALSE)

All current versions:

multinma dev ver 0.5.1.9007 loo ver 2.6.0 rstan ver 2.32.5 R version 4.3.2 (2023-10-31) macOS Sonoma 14.2.1