This is now the page with the big table
for older info
Calibration Checks: calibrationChecks.txt
1. Weird jump in the middle of galaxy 45 min. scan. (Note that these data haven't been fully flagged otherwise.) I think this is what Steven also saw: n3109-jump.jpg
. Note that the galaxy appears to get brighter in (some? all?) antennas in the lowest scan. It then appears to be calibrated out for the rest of them. I think the only thing we can do is flag. Which part do we want to flag? The higher part?
Flaring Phase Calibrator Issue:
The issue is that we see linear flaring in the calibrated phase of the last phase calibrator scan. This only occurs when the phase calibrator and the flux calibrator are the same (i.e., 1331+305), in cases with galactic HI interference where the calibrators are observed at different frequencies. In these cases, we derive SN tables in CALIB for each FREQID separately. We then copy all the SN tables to LINCOP, merge them, and run CLCAL on the merged SN tables. Normally, CLCAL does not interpolate between sources with different name, so in general, it does not interpolate between the flux and phase calibrator. In this case, however, it does (because they are the same). The phase solutions derived for the offset frequency are different. Therefore, the interpolation between the phase and flux calibrators has a large linear slope - it is not (as) flat as expected. This should not be a problem, because there is no galaxy data between the last phase calibrator scan and the last flux calibrator scan.
If you plot up the systemp temperature using varplt in miriad, they are all set to 60.. (varplt *.uv yaxis=systmp device=\xs). This is no good! -- AdrienneStilp
- 12 Aug 2008
Do we want to image our calibrators to make sure they look okay? It's a good check of the calibration, and the data should already be pretty well flagged, so it shouldn't be too many more steps. Do we want to set options=mfs in miriad's invert task if we do this? -- AdrienneStilp
- 06 Aug 2008
JO (19 Aug): No need to do this. In fact it only makes sense when you have another calibrator to be imaged. The 'calib' task assumes that the phase calibrator is a point source and self-calibrate it. So the phase calibrator will always
look like a nice point source because calib forces it to. Imaging the bandpass calibrator with the solution from the phase calibrator will be a better way to assess the quality but remember that the bandpass calibrator is not bracketed by the phase calib and that it is usually far away and the phase solutions won't be the most valid ones.
Let's put our calibrator models (for CALIB) on disk 1. That way the reduction scripts can keep the same getn 3, getn 4, getn 5 throughout. What will change is get2n, but that will have to be changed specifically for each galaxy anyway. I'll update the reduction script with this at some point. -- AdrienneStilp
- 31 Jul 2008
JO: Does that work on the different user IDs?
AMS: Different user IDs will start off with new catalog numbers - they know nothing about other user IDs. The problem is that when you start working on, say, Sextans A in C array after finishing Sextans A in B array, it will be (getn in the script) + 5 for the second obs. of the same galaxy. Does that make sense?
I've noticed that in most of the data sets I've looked at (ddo6, ngc4163, sexa) that there's a spike for an entire time on source in the galaxy data in the central part of the data set. I'm not sure if that's there for all the array configurations yet. It seems to be contained to certain antennas. -- AdrienneStilp
- 01 Aug 2008
JO: Is this just a single channel? This sounds pretty much like a correlator issue. We may need to flag this one channel. If it's more than a single channel, I'll look into it. And is it all EVLA
Current Reduction Recipe (as of Oct. 24, 2008):
: Recipe for NO Galactic HI
: Recipe for Galactic HI
In both the recipes, you can search and replace for various fill-in-the-blanks. For example, replace:
<phase cal> with the name of your phase cal,
<flux cal> with the name of your flux cal
<galaxy> with the name of your galaxy
<n> with the catalog number of the CH 0 file that AIPS reads in
<r> with the refant
The recipes have been.. mostly tested.
- 24 Oct 2008
Here's JO's version of the spreadsheet:
: version Oct 30 2008
- 19 Jul 2008