Closed wwiecek closed 8 months ago
Consider
loo <- loocv(schools) loo
(1) Why is standard error negative?! Probably because Danny multiplied this by -2 in the old implementation. Check if documentation of loocv() is clear to a casual reader.
loocv()
Now consider
loo2 <- loocv(schools, prior_hypermean = normal(10, 1)) loo_compare(loo, loo2)
(2) When printing loo we should also print one line explaining how the results should be interpreted.
loo
Check if documentation of loo_compare is clear ot a casual reader.
loo_compare
Consider
(1) Why is standard error negative?! Probably because Danny multiplied this by -2 in the old implementation. Check if documentation of
loocv()
is clear to a casual reader.Now consider
(2) When printing
loo
we should also print one line explaining how the results should be interpreted.Check if documentation of
loo_compare
is clear ot a casual reader.