So yes go ahead and leave the special case for empty assemblies; please just add a comment as to why it must be handled specially. Thanks!
Done in commit 779c026!
How do those differ? When there are no elements, the whole configuration vector space is the zero space...
Good point: in principle, a ConfigSubspace variant representing a full subspace…
I've dropped the eigenvalue logging code (in commit eafb133) and filed an issue (#106) to remind us to replace it.
Any reason not to go ahead and do that in this PR?
On second thought, we'd actually have to use a separate conditional for the unconstrained case, because the tangent space would be…
I'm fine with dropping the eigenvalue logging now and replacing it with a diagnostics panel plot later—either in a stand-alone pull request or as part of a pull request that revises `ConfigSubspa…
(a) Why do we now need an empty-assembly special case, when we didn't before? Or is it just for efficiency?
The realization triggering system on the main branch never tries to realize an…
What does this #-directive do? Not compile the next line if this is not a "dev" compile?
Correct, if you replace "not compile" with "only compile" (I think this was a typo, based on your next…
@glen wrote in StudioInfinity/dyna3#104 (comment):
OK, worked this time. Perhaps the first time I accidentally put a -0.25 into one of the A-C…
To help check whether we're following the same procedure: I'm setting distances in this order.
From
Huh: I don't get a line search failure. In fact, once I've set all the inversive distances except the A–C ones, I find that the A–C ones already have the desired values. This is consistent…