Andreas Borgen Longva

> > While this should have negligible overhead for larger matrices since the decomposition itself is `O(n^3)` > > It might have a large memory overhead if nalgebra doesn't have...

@timjb: singular values are now sorted, but IIRC, eigenvalues are not yet sorted (unless I missed a PR for this!).

Thanks to the work of @aarsenij in #1000, we now have an implementation for CSR matrices! :tada: Besides an implementation for CSC matrices, there is some room for improvement in...

Hi @p-e-w, thanks for your input here! First, let me preface this by saying that my opinion here is not in any way authoritative: I'm a heavy user of `nalgebra`...

> but in the specific case of `determinant()`, I think that the trait bound should be `Number` and not `ComplexField`. The problem here is that computing determinants of general integer...

> As far as additional maintenance burden is concerned, I honestly just don't see it. Formal matrix operations are trivial to implement, and the correct and future-proof trait bounds should...

@aujxn: ah, I had totally forgotten about the naming of "outer" iterators. Right, I suppose `Inner` is not too bad then :-) Yeah, in retrospect the naming of the outer...

Hi @oezgurmakkurt, this looks very cool! Great job :-) Question: is the benchmark against `matrixmultiply` single-threaded for both `matrixmultiply` and `blasoxide`, or is `blasoxide` using parallelization? Afaik, `matrixmultiply` is only...

That's really cool! Having a library competitive with OpenBLAS in pure Rust is amazing, because we can avoid all the painful issues surrounding compilation and dependency management. One issue with...

Well, as excited as I am about this, in the end this is all the decision of @sebcrozet, so perhaps wait to see if he agrees with what I wrote...