Open aarchiba opened 3 years ago
Can we put a hot fix for it?
This was triggered by a failure in the NANOGrav 15yr autoruns, which was reported as a bug: https://gitlab.nanograv.org/nano-time/timing_analysis/issues/67
That bug is now closed, suggesting that this is no longer a problem for autoruns.
I can confirm that this happens, and that incorporating the fitted par file into git resolves it.
The fitter actually works in this situation - I'm not sure why, I guess the fitting code treats unset parameters as having zero values, but the reporting fails. This could be fixed by the suggestions in #1124 or by special-casing a value of None in one more place.
And the fitter's
model_init
is: