Ilhan Polat
Ilhan Polat
Apart from the failing array API test, this is sort of ready. I'd appreciate if we don't squash merge it in case there are no comments.
@rkern Just out of curiosity, I have been using the NumPy C API for arrays to send things around to Lapack etc. For that, thanks to Cython, there is a...
But in those examples we don't have any explicit pass of a `Generator` hence my question. So something along the lines the user passes `rng` created with `rng=np.random.default_rng()`. Then do...
Ah OK thank you! I'll play around with it.
This one is ready to go in case there are no further comments.
I've gone through this again in the last two days with random inputs and did not notice any differences between the old fortran code. It turns out `diffsnorm` was broken...
@j-bowhay @h-vetinari We have two functions `scipy.linalg.interpolative.rand` and `scipy.linalg.interpolative.seed` deprecated with this PR. Rand is for generating random numbers, not sure why we had it in the first place, and...
Thanks @VegardFalmaar this is indeed something we really needed. I have been doing a similar thing for SLSQP. So this is good timing. One detail that we have to be...
> One always has the opportunity to use the same logging mechanism for all loggers by configuring the "root" logger since it is the parent of all other loggers. The...
> By default, that configuration on the root logger will cascade down to all of the named Loggers (unless if specifically overridden). This is what you do only if you...