Patrick Schultz
Patrick Schultz
For the normal, poisson, and chi-square distributions, we have sampling (`rand_norm`), density (`dnorm`), cumulative probability (`pnorm`), and quantile (`qnorm`) functions. For beta and gamma distributions we only have the sampling...
For example: ```scala // build.sc import mill._ import mill.scalalib._ object foo extends ScalaModule { def scalaVersion = "2.12.15" } // foo/src/Foo.scala package foo object Foo { val value = "hello"...
Slightly tweak the `Bindings` class to store a single array of (name, type) pairs, with the various environments referring back to that array. This simplifies passes that want to create...
Greatly simplifies the `NormalizeNames` pass to be logically independent of the binding structure. This leaves the old implementation, and asserts that they agree. Will delete the old implementation in a...
Makes some further progress on simplifying the `PruneDeadFields` pass, with the primary goal of decoupling it from the details of the binding structure. The primary change is to `memoizeValueIR`. Before,...
### What happened? A user reported this bug on discuss: https://discuss.hail.is/t/illegalargumentexception/3808. This looks like it may be related to #14420. ### Version 0.2.130 ### Relevant log output ```shell In [13]:...
We got a [request in zulip](https://hail.zulipchat.com/#narrow/stream/127634-Feature-Requests/topic/String.20find) to add a function that works like the python [`str.find()`](https://docs.python.org/3/library/stdtypes.html#str.find) function. This should be straightforward. We just need to register a function in `StringFunctions`...
After discussing with @ehigham, we decided to try a lighter-weight RFC process, where an RFC is simply a PR for a new dev doc. Discussion about the design can happen...
If a user writes a map over a local array or stream value, which they know is large, they should be able to force scratch memory to be cleared after...