Provides same interface to studentized range distribution functions ([p/q]tukey) #66
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
pdf code was written originally by user "Doomphoenix Qxz" on discourse.julialang.org
simonbyrne
reviewed
Jan 28, 2019
simonbyrne
reviewed
Jan 28, 2019
Vast improvement in performance.
simonbyrne
reviewed
Jan 29, 2019
simonbyrne
approved these changes
Feb 1, 2019
matbesancon
approved these changes
Feb 2, 2019
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
To support an implementation of proposed
StudentizedRange()in Distributions.jl.The interface of the Rmath library is a bit different for the tukey functions than the others, with a parameter (nranges) set to 1 which isn't really explained by the R documentation. Thanks to the Rmath.jl package which showed me the right argument ordering.
All are a bit slower than the other distributions, ~1 order of magnitude.
Note: there is a case where the underlying ptukey and qtukey functions diverge with higher df/k and values closer to 0. This is documented as a broken test and with a special case in testing to avoid those values.