allow predicting from a single observation #653
Conversation
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Codecov ReportBase: 94.23% // Head: 94.23% // No change to project coverage 👍
Additional details and impacted files@@ Coverage Diff @@
## main #653 +/- ##
=======================================
Coverage 94.23% 94.23%
=======================================
Files 29 29
Lines 2795 2795
=======================================
Hits 2634 2634
Misses 161 161
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
There was a problem hiding this comment.
given that the Grouping struct itself doesn't keep track of the levels, I think all the validation stuff can just be skipped altogether. also, would it break anything to just remove the termnames/contrasts_matrix methods altogether? I think the only place they get used is in these constructors...
Co-authored-by: Dave Kleinschmidt <dave.f.kleinschmidt@gmail.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
src/grouping.jl
Outdated
| StatsModels.contrasts_matrix(::Grouping, baseind, n) = zeros(0, 0) | ||
| StatsModels.termnames(::Grouping, levels::AbstractVector, baseind::Integer) = levels | ||
| # StatsModels.contrasts_matrix(::Grouping, baseind, n) = error("Grouping terms don't have associated contrasts") | ||
| # StatsModels.termnames(::Grouping, levels::AbstractVector, baseind::Integer) = levels |
There was a problem hiding this comment.
probably better to just clean up this cruft, OR leave the methods as they were before with a note that they're probably not needed since we've added a direct constructor for ContrastsMatrix
Co-authored-by: Dave Kleinschmidt <dave.f.kleinschmidt@gmail.com>
closes #625
docs/NEWS-update.jlto update the cross-references.