Add normal and exponential distributions to core::rand #6098
Conversation
This file contains hidden or 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
|
@huonw Needs rebasing :-( |
This replaces the wrapper around the runtime RNG with a pure Rust implementation of the same algorithm. This is faster (up to 5x), and is hopefully safer. There is still much room for optimisation: testing by summing 100,000,000 random `u32`s indicates this is about 40-50% slower than the pure C implementation (running as standalone executable, not in the runtime).
Sample from the normal and exponential distributions using the Ziggurat algorithm.
|
@catamorphism done! |
flip1995
pushed a commit
to flip1995/rust
that referenced
this pull request
Oct 9, 2020
Downgrade interior_mutable_const lints to warn by default This change updates the two lints in the file non_copy_const.rs to be warn by default rather than deny by default. It also updates the known problems for declare_interior_mutable_const to mention some issues that are affected by the lints. This is a repeat pull request since I botched the first one (rust-lang#6012). Apart from my messing up the commits of that one, I also had a problem where the stderr of the tests didn't change despite me changing both lints to warn by default. Is this normal behaviour for some lints or do I need to adjust the tests to accommodate the change? fixes rust-lang#5863 changelog: none
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Outdated suggestions cannot be applied.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.
(This is only a94c86e, the other two are from #6073.)
Very basic support for sampling from N(0,1) and Exp(1). This is basically my experimentations with the Ziggurat algorithm, and will certainly get reorganised when Lib-rand gets finished.
No tests or anything, other than generating 1 million numbers from each distribution, loading it into R and plotting and performing some basic tests (which passed) i.e. Pearson's chi-squared for the Exp(1) values, and Shapiro-Wilks for the N(0,1) values.
(These will be tested better when the lib rand changes linked above start to happen.)
(Also, it's not hugely relevant, but I noticed that this is about double the speed of R for summing 10000000 samples from the normal distribution, even though the current random-float code is subobtimal.)
No need to merge this if it seems like it's being too hasty.