- Notifications
You must be signed in to change notification settings - Fork68
Open
Description
I think the output for the ordbeta family is a bit confusing. In particular it just lists the "lower cutoff estimate" but then lists two different numbers. The second one appears to be the upper cutoff estimate.
Example:
n <- 100x <- rnorm(n=n)y <- plogis(x + rnorm(sd=0.5, n=n))y[y>0.9] <- 1y[y<0.1] <- 0mod <- glmmTMB(y~x, family=ordbeta)The output when you just callmod:
Formula: y ~ x AIC BIC logLik -2*log(L) df.resid -139.07483 -126.04898 74.53742 -149.07483 95 Number of obs: 100Dispersion parameter for ordbeta family (): 20 lower cutoff estimate: 0.0128, 0.993 Fixed Effects:Conditional model:(Intercept) x -0.02224 0.95327It would also be nice if the cutoffs were available in calls tosummary() Also is it possible to get the standard errors around them?
Metadata
Metadata
Assignees
Labels
No labels