Call slot for a model fit with glmer.nb

[pitakakariki]

example(glmer.nb, run.dontrun=TRUE)
getCall(m.nb)

This gives us glmer(formula = y ~ f1 * f2 + (1 | g), data = dd, family = negative.binomial(theta = 0.4797700780901)), rather than glmer.nb(y ~ f1*f2 + (1|g), data=dd, verbose=TRUE) which we might expect.

Would anyone mind if I changed this?

[pitakakariki]

Ah, looks like this would cause downstream problems for the effects package?

[bbolker]

how did you figure that out that fast? Did you run all the downstream tests? (I'm looking for a more reliable, less annoying platform for doing downstream testing ...)

[pitakakariki]

It failed three of the testthat expects, and you'd left a helpful comment about the effects package!

[pitakakariki]

Looks like there's something planned for reverse dependencies at r-hub: https://github.com/r-hub/proposal#reverse-dependency-checks-week-19