Peter

What kind of value would your dataframe contain? number / string / datetime?

Yes, you can always use `PyCall.jl` or `PythonCall.jl` to call those python function directly. The purpose of this package (`Pickle.jl`) is to provide a pure Julia implementation so that we...

@CarloLucibello This is great! but I would doubt if it's possible to become a julia-only package. I took investigation long ago when making HuggingFaceApi.jl, and realized that the way they...

I don't really have any concrete idea right now, but it seems worth to split out that rule part. For example, if someone prefer the older `IdDict` based update implementation,...

> FWIW, there is an IdDict implementation ... It's a lot simpler than all this tree-walking stuff, I have to say... I would say that's the reason why I think...

Ok, that's fair. > Not sure how we got higher derivatives, IIRC they aren't used by any rules here & aren't even tested, no idea if the design is right....

I see. BTW, I turn into using `findfirst` as a replacement and it doesn't seem to be forwarded to `_mapreduce`. I'm curious about would these "find" function also benefit from...

I use `SizedVector` merely for having the length value as type parameter and avoiding `push!`/`pop!` operations at the surface level. Maybe there is a better choice for this need?

@mschauer but most of the api called `something` beforehand, so that efficient algorithm never has the chance to run and even we provide the `mean` it will still calculate it...

@mschauer It's not only implemented for `dims=:` but also only for iterable. For `AbstractArray` type, the problem I mentioned above is not solved. The `mean` should not be calculated twice....