The GW150914 parfile uses Llama’s z symmetry (bitant) to reduce cost. However there is no check in the script that enforces this nor is it (at all) apparent from the Python code that z symmetry is used. The most obvious issue is that the script seems to offer the user to change all 3 components of the reduced spin chi while for z-symmetry only the z component of spin may no non-zero (spin being an axial vector).

Pull request

https://bitbucket.org/einsteintoolkit/einsteinexamples/pull-requests/3/gw150914rpar-make-z-symmetry-usage

removes the x,y components of chi from the script and adds an explicit check for symmetry violating ID parameters to the script.

--
Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2486/make-z-symmetry-usage-explicit-in-gw150914