bglmer and residual prior
bachlaw opened this issue · comments
bachlaw commented
Is the resid.prior
option intended to only be supported for blmer
models? I was trying to set a gamma (1,1) prior on the residual variance (essentially, a unit exponential prior) which of course tends to be very useful, but bglmer
reports that resid.prior
is not an available option. It does appear to be a viable option for blmer
however.
Vincent Dorie commented
When I wrote the package, glmer (and hence bglmer) models didn't have a
residual variance parameter. If any families have been added since then, I
can look into supporting them. What kind of response family are you trying
to fit?
…On Sat, Apr 14, 2018 at 3:30 PM, bachlaw ***@***.***> wrote:
Is the resid.prior option intended to only be supported for blmer models?
I was trying to set a gamma (1,1) prior on the residual variance
(essentially, a unit exponential prior) which of course tends to be very
useful, but bglmer reports that resid.prior is not an available option.
It does appear to be a viable option for blmer however.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#5>, or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAnX8NcsZiC10-cBzqwFW4ahdOpshZWpks5tok5lgaJpZM4TVMde>
.
bachlaw commented
Come to think of it, they still might not. Rats. In terms of the model, just a logistical binomial at the moment.