Phillip Alday
Phillip Alday
> A higher R^2 indicates a better model No, even assuming that there is a clear definition of R^2 for a given model (which there isn't for mixed models). https://www.johnmyleswhite.com/notebook/2016/07/23/why-im-not-a-fan-of-r-squared/
> But it still has a kernel of truth at its core: lower variance means better predictions. Again, no! For one thing, R2 captures a particular measure of variance (relative...
I think @ararslan had an implementation somewhere -- oh yeah as part of his work on Cook's Distance for GLM: #510 @HiramTheHero your proposal is mathematically correct but generally computing...
closing in favor of https://github.com/JuliaStats/StatsAPI.jl/pull/26
I think we could add a kwarg `validate_record_count=true` -- really highlights why this exists. There are a number of good and bad reasons why the record count in the header...
@nalimilan I started working on this again while other stuff was compiling and running today, but the tests -- even on `master` -- have both failures and errors on Julia...
This feels an old issue where the Julia links to a different version of some C/C++ standard libs than R does. [In the past](https://github.com/JuliaInterop/RCall.jl/issues/268), you had to fix this by...
Can you provide an MWE of how you reached this error?
Closing as very stale at this point
> It may be better to return a `Vector{DictTable}` rather than a `NamedTuple{DictTable}` because the expression for the grouping factor is the name of the keys in the `DictTable`. >...