Conversation
jurjen93
changed the title
| f'ac{ac_count}.parmdb={applycal_h5}', | ||
| f'ac{ac_count}.correction={corr}'] | ||
| f'ac{ac_count}.correction={corr}', | ||
| f'ac{ac_count}.missingantennabehavior=flag'] |
There was a problem hiding this comment.
A clarifying question: what drives the choice for this parameter? Since the delay solutions are pre-applied for the DD calibration one way or another, I would not expect a mismatch in stations at this point. Are there cases were stations are removed from the solutions during the DD calibration?
There was a problem hiding this comment.
Yes, I had a case where flagged stations were removed.
So, instead of ending up in an error, and having to add these back yourself manually, its easier to flag these during the applycal
There was a problem hiding this comment.
I don't see any other use of missingantennabehavior in the pipeline at the moment. Doesn't this imply that all stations behaved ok during the delay calibration and that this is strange? I wonder in that case whether this is overlooked earlier in the pipeline, or what would cause the DD solutions to fail so catastrophically that stations needed removing in that case.
There was a problem hiding this comment.
I guess it comes from: https://github.com/LOFAR-VLBI/pilot/blob/master/steps/make_concat_parsets.cwl#L46
This option removes fully flagged stations from the MS. This is a handy feature because it saves some storage space for large datasets.
There was a problem hiding this comment.
Maybe this should be earlier done in the pipeline though (not only during dd-calibration and the facet subtraction), because it indeed comes from flagging earlier in the pipeline.
There was a problem hiding this comment.
In my specific case I dont have access to the delay calibration solution plots or the solutions themselves to see what happened (those were generated by James and I dont find them)
2 participants
Uh oh!
There was an error while loading. Please reload this page.