This confuses me: isn't the point of DAMD-11 to avoid publishing the one-time hack redFiberTraces.fits? If we get any formal wavelength solution persistence then we can forget that we ever needed this file.

Until DAMD-11 is implemented we still want the redFiberTraces.fits to make the tests pass again. As the format has changed compared to the old version of the file some code changes are required for which a ticket is needed.

Ready for review once PIPE2D-119 is merged into master. Tests in Spectra.py pass again.

Changes are in drp_stella and obs_pfs tickets/PIPE2D-124

Per our meeting of 2017-01-09, aritter will check if this ticket is actually still necessary. Since we expect that this work to be rendered obsolete by DAMD-11, this will be marked as "won't fix" unless it's actively breaking tests in the short term.

It turns out that the old file RedFiberPixels.fits.gz is sufficient since the geometry bug was fixed in Craig's simulations. However I recommend adopting the remaining changes in drp_stella tickets/PIPE2D-124 as they are improvements to the code. Part of the changes undo changes we made when the new test data containing the geometry bug in the simulations were introduced. Another part removes un-used imports and variables. We could rename the issue and submit it to review, which should be a quick one...

Given the above, we agreed at our meeting of 2017-01-18 that this ticket should be closed as "won't fix", and we need three new tickets to take care of the work which currently exists on the PIPE2D-124 ticket branch (which is worth saving).

aritter will file those tickets then mark this as "won't fix".

Removing the `yCenters` parameter from `assignITrace` and `findITrace` is done as part of PIPE2D-138.
Removing unused import from testDrp.py will be done in PIPE2D-69.
Filed ticket PIPE2D-149 to replace `reduceArcRefSpec.py` with `reduceArc.py` in the quick-start guide.