Does GPAVER work when using nfbin in GPCAL?

[shane]

screen-capture.png (108.51 KiB) Viewed 38601 times

Hi,

I go through my calibration and get an nice result (Fig.1) for Stokes I. Using gpcal with nfbin=5.

gpcal vis=$cal2 tol=0.0001 interval=0.1 refant=3 nfbin=5 options=qusolve

Then I run gpaver and it creates a nice symmetric mess (Fig.2).

gpaver vis=$cal2

I've tried options=scalar and vector and changing the interval.
The calibrator data is ~4mins long and taken ~every hour.

So my question is does gpaver work when using nfbin in gpcal?

If it does, I have no idea where I'm going wrong!

cheers,
Shane

[ste616]

Hi Shane,

Thanks for the report. This looks like a Miriad bug since the gpaver task is not currently able to do averaging on the frequency-dependent gains, and the frequency-independent gains should not be used during plotting if the frequency-dependent gain tables are present. We will look into this issue as soon as possible.

On a slightly different topic: this is not the correct mode for gpaver use. You should only average the gains after they have been copied to the science source, so as to keep the full gain solution available from the gain calibrator. If you are doing the gpaver on your gain calibrator only to show the error then please ignore my concern!

[shane]

Hi Jamie,

Thanks for the info and the advice on the correct use of gpaver!

Cheers,
Shane

[Mark.Wieringa]

Hi Shane,

gpaver has now been updated to deal with frequency binned gain solutions.
The new version should be available everywhere tomorrow.
Thanks for alerting us to the problem.

Cheers,

Mark

[shane]

Great!
Thanks Mark.

Shane