Change CER flavor from CERQUICK to CERAUTO

[odstrcilt]

No description provided.

[AreWeDreaming]

Why? I don't use CER data regularly so I don't know about the flavors

[odstrcilt]

CERQUICK - Fast analysis available a few minutes after discharge. The primary purpose is for analysis in the control room. It uses several simplifications. It should not be used for detailed analysis.
CERAUTO - Overnight analysis includes lines to fit the passive component of the charge exchange line better. It also includes a quality check to filter out poor measurements.
CERFIT - Manually done CER analysis and usually based on background subtraction, and traditionally required for the analysis of non-carbon impurities. Some of these impurities are recently done by CERQUICK and CERAUTO, too, but the quality is generally low. Although CERFIT accounts for ~0.1% of all shots, it represents a significant fraction of thoroughly analysed, actually published discharges.

OMFIT profiles selects just the best available edition. But this is not a good idea for automatic analysis, because there are discharges where CERFIT analysis is done only for the impurity and not for the carbon. In this case, both editions needs to be loaded and merged. And then there are even some discharges where CERAUTO and CERFIT editions contains a fit of different CX lines for the same channel.

[smithsp]

@odstrcilt We have tried to get Colin Chrystal to allow us to run CERAUTO between shots. He hasn't wanted to put in the effort, but we (Sean F and I) think that there is sufficient computing power to run CERAUTO between shots. If you could help advocate for this, we could do away with CERQUICK.

[smithsp]

@torrinba This needs the mongo fix, it seems.

[odstrcilt]

If you are planning to use it shortly after the discharge, before CEAUTO is availible, that the best solution would be to pick the highest CER flavour which includes carbon data.
OMAS current code is iteration over range(1,100) channels (instead using only existing channels),
includes the non-carbon impurity channels (this is often undesired behaviour since the impurity label is not saved). Also it is loading TEMP instead of TEMPC ( Zeeman splitting and fine structure corrected temperature).
I can write a script which selects the best edition from MDS+, however I'm not sure how and where to include in the OMAS.

[github-actions]

This PR has not seen any activity in the past 60 days. It is now marked as stale and will be closed in 7 days if no further activity is registered.