Erik Welch

Understood, thanks for explaining the benefits and costs. If these are ever available--even optionally--I would opt-in to include them. I'm not so sensitive to binary size, but I know this...

Thanks. I guess I don't understand what's unwieldly about this other than the risk that some day you'll have both `GxB` and `GrB` versions. Adding the inner product adds *one*...

But certainly take your time and do what you feel comfortable with. You provided workarounds, so this is a non-blocking issue. Heh, I wrote this issue out of frustration at...

Great. Dot product really does come up pretty often for me. I'm computing global clustering coefficient right now, and the denominator is naturally obtained from a dot product of vectors....

For `GxB`, I think it would be sufficient of you only added outer product and kron for BinaryOp just like how you only had `GxB_kron` for matrices with BinaryOp. Hence,...

I decided I like the names `inner` and `outer`, which is what I'm adding to `grblas` using your casting trick (works great!).

Note that this behavior was added to pandas 0.23: https://pandas.pydata.org/docs/whatsnew/v0.23.0.html#whatsnew-0230-enhancements-merge-on-columns-and-levels and was added to Dask here: https://github.com/dask/dask/issues/2950 https://github.com/dask/dask/pull/2960 This is an important optimization for `dask.dataframe` and `dask_cudf` to avoid full...

Thanks. I can reproduce. This is actually an error in `dask`. Here is an example that goes through a similar code path that give the same error: ```python import dask.dataframe...

Fixing in https://github.com/dask/dask/pull/9485

This issue can be closed. This is fixed in dask version 2022.9.1, which was released on September 19.