Kevin Kuo
Kevin Kuo
What's the latest on this issue? I'm considering adopting mleap for [sparklyr](https://github.com/rstudio/sparklyr) deployment but this would be a blocker since most of the canned pipelines in the package utilize `RFormula`.
+1 for support of Spark `ImputerModel`, makes exporting existing pipelines much easier.
We'll be taking a closer look at this when we add mleap support to the mlflow R interface. Hopefully it's just a simple configuration issue but we'll see!
I just hit this, too. My use case is I want to include an organization name as an author, which I can't figure out how to do in distill right...
@emitanaka FYI in case you're using radix, the organization author issue has been fixed on master.
@colah could you PTAL?
Wanted to point out that, in general, providing a Scala API for a Spark extension also means that R users get access to it (via sparklyr.)
Not implemented yet but will look into adding it in the next release.
The cleanest way to do is probably using a temp dir. We also want to compile against the right mleap depending on the spark version. Let's kick this can down...
+1 that the streaming features should be exposed at the level consistent with existing functions, and sugar can be built on top of them later, either in this package (with...