Max Veit

One at a time: > I feel it would be beneficial to separate file saving and computation functionality, e.g. `calculate_and_sparsify`. Sure, although some of that was intended as diagnostic information...

> Would just add these 4 lines in this branch, seems the simplest solution. The lammps integration PR will be very likely thrown away and split into smaller PRs as...

Ok, that's weird, what was the actual error you got? Was it coming from `gaptools._get_energy_baseline()` or from somewhere in the `KRR` class? It seems the current `KRR` class is very...

@ceriottm instead of messing with the notebooks when they really haven't changed in content, perhaps you could instead just do a `git restore .ipynb`? It would also give a much...

Ok, this is almost ready for review -- it has all the functionality I was planning to add here, now just needs some testing. The main major change since the...

I can confirm that downgrading Sphinx from v4.1.0 to v4.0.1 (and changing nothing else) fixes the doc build. I think we should report the issue to the Sphinx developers --...

I would say it's not a priority. We used row-major ordering where we needed to pay explicit attention to the memory ordering (e.g. where we had to compute some indices...

This is not, in fact, fixed by #403 for me: ``` [ 58%] Linking CXX executable forwarding_of_property_requests ld: warning: ignoring file CMakeFiles/forwarding_of_property_requests.dir/forwarding_of_property_requests.cc.o, building for macOS-x86_64 but attempting to link with...

Fixed now, seems the issue was that it was linking to Intel libraries that were basically just copied over during the migration. The best remedy for this situation is either...

Just to note this is still an issue, and now it interacts with `src/rascal/math/kvec_generator.cc` spewing all sorts of incomprehensible garbage about a mysterious apparently-internal Eigen variable being outside of array...