Alex Arslan

Yes?

Thanks, I appreciate it :smile:

> Unfortunately this cannot live in StatsAPI as it calls `StatsBase.cov2cor!`. `cov2cor!` is actually from Statistics (`Statistics.cov2cor! === StatsBase.cov2cor!`), so AFAICT it should be fine in StatsAPI.

Are you envisioning something like ```julia std(x; correction=-1, mean=mean(x)) = sum(xi -> (xi - mean)^2, x) / (n + correction) ``` ? Or are you thinking more like the ability...

I'm not sure how these things work but for what it's worth the repo license lists the following as the copyright holders: > Copyright (c) 2012-2016: Dahua Lin, Simon Byrne,...

We could also use https://github.com/c42f/FastClosures.jl, which I believe does the hack I described above, but is a registered, maintained(?) package. It does not support `do` blocks though, which I have...

> but instead being there own singleton functors, that are a lot like closures. That's interesting. So they would work kind of like how functions work in Julia, where they...

As an aside, the `@eval` shouldn't be necessary; instead, the returned expression could be `:toplevel`.

I'm not sure it makes sense for that to be in ChainRules; as far as I understand, checkpointing is fairly specific to the underlying implementation. Take for example Will's Nabla...

I noticed last night that we'd need to be careful about which argument is the one we're checking. Consider e.g. BLAS functions, which often have the differentiable argument last.