Phillip Alday

Include CIs for Generalized Eta-Squared

3

Including Generalized Eta Squared in the output for `ezANOVA` is really great for getting people to start looking at effect sizes and not just p-values. But we should also consider...

`merge` and `unique` kwargs with Legend on user specified plots + legends

4

The `merge` and `unique` kwargs don't work for the `Legend` method where the user specifies an array of plots and labels (instead of collecting the plots and labels from an...

`ftest` and `-0.0`

3

Currently, `ftest` uses the unadjusted R2, which is constrained by definition to be on [0, 1]. As it stands in the tests and doctests, we have several cases where the...

See #415 - [ ] specific method for Cholesky solver `LinearModel` - [ ] specific method for QR solver `LinearModel` - [ ] general fall back method using QR solver...

methods for `StatsBase.leverage`

3

Now that we have Cook's Distance, we can define this rather straightforwardly.

API for VIF

10

Variance Inflation Factor -- if we ever move `StatisticalModel` and `RegressionModel` to StatsModels, then this should go there, but until then, I'm guessing this is the best place. Next step...

Preserve eltype where possible for moments

6

Fixes #680.

xref: https://github.com/JuliaCloud/AWSS3.jl/issues/263 The fix seems to be change `::Type{Union{Missing, T}}` to `::Type{

use `axes()` and `eachindex()` to better support arbitrary `AbstractArray`

4

Starts the process of fixing #1265. Other things that I saw while doing this: - There are times where there is a constraint on the eltype but not on the...

xref: https://github.com/JuliaStats/StatsModels.jl/pull/261