Luca Palmieri

I'd say to use `clone` @munckymagik

Given that the `skipnan` functionality is not available in `ndarray` itself I'd say that this would be a good addition for convenience :+1:

There is a PR open https://github.com/rust-lang-nursery/rust-cookbook/pull/560

Cosine similarity would fit quite well in the scope - we already have a bunch of distance functions, so we can definitely expand the collection. What else do you have...

https://docs.rs/ndarray-stats/0.3.0/ndarray_stats/trait.EntropyExt.html#tymethod.kl_divergence :grin:

I agree with the conclusions of your analysis - waiting for https://github.com/rust-lang/rust/issues/31844 to get implemented is probably the best course of action for now. Unfortunately it is not going to...

Sure, let me do some more investigation :+1:

I haven't been very active recently, but I am still very much following this discussion - I am happy to contribute [linfa](https://github.com/LukeMathWalker/linfa) to a working group organisation if it does...

Re: communication channels - no strong preference either way.

I have two issues with struct update syntax: - it's not "obvious", as in something you would do if you were given an API that allows it. Most people would...