Evgeni Burovski
Evgeni Burovski
Agreed on both counts: the `dask.array.asarray(copy=None)` is a conscious break of the version policy so that the dask wrapper layer is forward compatible; there are no views at all in...
> I can open a PR, since the fix appears to be trivial. Superb, please do! That fast path was meant to be temporary until we find a better general...
Basically, torch `uintN` support is very incomplete: https://github.com/data-apis/array-api-compat/pull/253
Cross-ref a previous discussion : https://github.com/scipy/scipy/pull/6591
The smoke-docs failure is you need to make the string raw, `r'Convergence...'` ``` > plt.title('Convergence for $f(x, y, z) = \exp(x)$') ``` A +1 from me otherwise (whether this should...
No idea why it did not show previously on CI, was it masked by something else maybe? The relevant incantation is (with or without -v and with -s having the...
@mdhaber Just to double-check: the 4kLOC `_levedev.py` file records quite a bit of history (great!) and the last step is "this python code has been translated from Matlab" --- you...
Merged! Thanks @mdhaber, all.
> We could maybe use std::complex, but doing that opens yet another can of worms. Mind expanding on what is this can of worms?
We run modified doctesting of public objects on CI via `$ python dev.py smoke-docs` . Since `_tanhsinh` is not public, it is not doctested though. You may run doctests manually...