Use 100 km/s as the "exclusion" zone for identifying "good minima" in the redshift scans for stars, versus 1000 km/s (which remains the same for galaxy and quasar scans).
With this branch, using the example object discussed in #276 and #277, we now get the following minima with the PCA/Jura templates:
The main point here is that the best-fitting STAR redshift (with either set of templates) is now z=0.0015657 vs z=-0.001475 previously (i.e., with main). (Also note the much smaller delta-chi2 value between the STAR minima.)
In @sbailey's beautifully labeled graphic, the new redshift corresponds to the better / best "first" (i.e., non-bad) redshift minimum:
So I think this PR is non-controversial and should be good-to-go.
coverage: 39.505% (-0.02%) from 39.528%
when pulling 19c7e84c44877e204daaa556307c1578e1c92a5d on max-velo-diff-star
into 04653df87a123c920446bbf4380f68bfdf190457 on main.
Simple fix for #277.
Use 100 km/s as the "exclusion" zone for identifying "good minima" in the redshift scans for stars, versus 1000 km/s (which remains the same for galaxy and quasar scans).
With this branch, using the example object discussed in #276 and #277, we now get the following minima with the PCA/Jura templates:
And the following minima with an early version of the NMF templates:
The main point here is that the best-fitting STAR redshift (with either set of templates) is now
z=0.0015657
vsz=-0.001475
previously (i.e., withmain
). (Also note the much smaller delta-chi2 value between the STAR minima.)In @sbailey's beautifully labeled graphic, the new redshift corresponds to the better / best "first" (i.e., non-bad) redshift minimum:
So I think this PR is non-controversial and should be good-to-go.