Michael Abbott

Results 1143 comments of Michael Abbott

Re names, first English: If you want to "stack" some papers, or some hay, then you don't put the pieces together on their long dimensions as that would fall over....

I'm sure I've suggested this `max`/`maximum` analogy too somewhere. Arguments against it, and in favour of a new verb, are: * One, if M is a matrix, this is the...

Thanks, I hadn't seen that iterator approach. Perhaps something like that could now be written as mixture of `Iterators.flatten` (same iteration) and `Iterators.product` (same shape, given the first element). But...

Indeed, that's pretty fast. (Notice we're penalising `reduce(hcat, xs)` by collecting twice, my mistake.) The analogy to `Iterators.flatten` is perhaps a useful way to think about this. If it collects...

Now that we have #32310, one question here is whether this should do `stack(eachcol(x)) === x`. At present this PR always makes a new array, but it could be permitted...

Bump? These comments https://github.com/JuliaLang/julia/pull/43334#pullrequestreview-835887585 are still what I think needs feedback.

See [here above](https://github.com/JuliaLang/julia/pull/43334#issuecomment-986135890) for my argument against `cat`->`concatenate` being like `max`->`maximum`. One un-`cat`-like behaviour right now is the inverse of `eachrow`: ```julia julia> stack([[1, 2, 3], [4, 5, 6]]; dims=1)...

I appreciate the interest here, but this is a long thread already. Can I suggest that an unstructured discussion of what's generally going on might be better on Zulip/Slack or...

Re names, "stack" vs "slice" are plausibly a pair, right? The weird part is "each", which (IIRC) was because these were originally iterators. One reason for the view asymmetry is...