Aleksey Khudyakov
Aleksey Khudyakov
My bet is naive error estimation. It always fails somewhere in parameter space. I'll check it out
Playing a bit binomial distribution shows that yes. It's a bug. Ironically test failure happens in region where algorithm performs tolerable.
It turns out main contribution to error comes from binomial coefficient.
I took second look. For probability main contribution to error is term `(1-p)^(n-k)` . It's possible to replace it with `exp(log1p(-p)*(n-k))` but it isn't clear win for small `(1-p)`. I...
I mark this issue "help-wanted". Contributions are welcome
I don't understand what this test should do. Function with provided signature could only compute `PValue` from `Int` and doesn't seems to be useful
This is indeed problem in partialSort and GHC. Check GHC [bug #5539](http://hackage.haskell.org/trac/ghc/ticket/5539) for details. I think it make sense to keep this bug open for information purposes
Vector does lack high level documentation. And unless someone steps up and write it it's going to stay this way. So I mark this issue as help-wanted
It would require changes in stream fusion framework. Currently streams use Monad for sequencing of effects
There's no explicit policy about that. This I believe an omission but it also bring question of how much is guaranteed? Sequence of random Word32 produced by generator is very...