Hello all,
Should already be fixed in the momentum_constraint branch (also fixes some other issues concerning matter sources).
Do you think we could make that branch the default branch at some point? I don't like this diverging so much.
Works for me as long as some point is after the NRDA. It is somewhat tested ie. solving for a TOV at rest gives similar results than using TOVSolver alone. Solving for the momentum constraint still gives results that are not useful (terrible results due to taking second derivatives). This can be somewhat mitigated by having non-zero center_offset[0] and [1] (ie move the punctures off the x-axis).
Probably best if someone other than me looks at the code as well before we make it the default (vacuum is backward compatible, matter sources are not).
Yours, Roland