Page 1 of 1

selfcal from model generated by mossdi

PostPosted: Thu Sep 10, 2015 5:37 pm
by luke
Hello,

I am interested in appling selfcal (http://www.atnf.csiro.au/computing/software/miriad/doc/selfcal.html) on a mosaic of 40 pointings. However, the clean component model I generate using mossdi will not work with selfcal. I get the selfcal output

Code: Select all
Calculating the model for data.ic
### Fatal Error:  Expecting only a single pointing, but got several

At first I thought this was a problem with the uv data having multiple pointings, but I do have 'options=mfs, mosaic'. Also the output clearly states the error is related to generating the model from the clean component file.

Since I would like to perform selfcal on a mosaic, are there any suggestions for the best way to do this?

It looks like I could use the task demos http://www.atnf.csiro.au/computing/software/miriad/doc/demos.html). This breaks the model into separate pointings, and reverses a primary beam correction. Separating the model into individual pointings could solve the problem. However, the dirty map mosaic was generated using invert for a joint deconvolution, so I assume it does not have a primary beam correction applied before generating the components with mossdi.

Re: selfcal from model generated by mossdi

PostPosted: Fri Sep 11, 2015 9:33 am
by Mark.Wieringa
Hi Luke,

I think you've answered your own question - you use demos to split up the mossdi model into separate pointing and feed those into selfcal together with the visibility file containing all the pointings. When you mosaic with invert it does use the primary beam correction to combine the pointings into a single image, so using demos returns the model components back to the uncorrected flux scale that is present in the visibilities.

Cheers,

Mark

Re: selfcal from model generated by mossdi

PostPosted: Fri Sep 11, 2015 12:14 pm
by luke
Hi Mark,

It is good to know I am on the right track. It is also good to know that invert does do a primary beam correction for the mosaic option, I always assumed it didn't.

Luke