Performance fails on rstanarm non "sampling" models

easystats / performance

:muscle: Models' quality and performance metrics (R2, ICC, LOO, AIC, BF, ...)

https://easystats.github.io/performance/

GNU General Public License v3.0

1.03k stars 94 forks source link

Performance fails on rstanarm non "sampling" models #5

Closed DominiqueMakowski closed 5 years ago

DominiqueMakowski commented 5 years ago

To be fixed once algorithm detection is implemented.

strengejacke commented 5 years ago

Do you have a reprex?

strengejacke commented 5 years ago

bump

I'm not sure what you mean by "non sampling" models.

DominiqueMakowski commented 5 years ago

sorry about that! I meant when algorithm is optimizing, meanfield or fullrank

strengejacke commented 5 years ago

Ok, and where does which function fail?

DominiqueMakowski commented 5 years ago

All I think :)

performance::model_performance(insight::download_model("stanreg_meanfield_lm_1"))
performance::model_performance(insight::download_model("stanreg_fullrank_lm_1"))
performance::model_performance(rstanarm::stan_glm(Sepal.Length ~ Sepal.Width, data=iris, algorithm = "optimizing"))

Because of LOO

strengejacke commented 5 years ago

Ok, finally I got it what you meant... :-)