Cube imaging- No visibility to map

[saracmbr]

Hi,

I've been trying to make a cube image of different stokes parameters but I face the error No Visibility to map.
There is no problem with options=mfs. There is no RFI or phase error in the image field(MFS).
The observation is at 16cm. I just tried a pre-CABB dataset (1.4GHz) and I get the same error.
Can anybody help me to find the problem.

Cheers,
Sara

[Mark.Wieringa]

Hi Sara,

it looks like there is some selection happening which results in there being no visibilities left.
E.g., you could be asking to image polarizations that are not present in the data, or the line parameter could be selecting flagged channels only.

Try a prthd in=visfile to see what is present in your visibility file. You can post the results here together with your invert inputs if you are still scratching your head after that..

Cheers,

Mark

[ste616]

Hi Sara,

The most common cause of this problem is due to the incorrect setting of the 'slop' parameter in invert.

By default, invert requires that every spectral channel (in the range that you are trying to image) be good before it will use that visibility. This is almost never the case, and certainly not in the 16cm band.

To allow invert to use visibilities that have flagged channels, set the 'slop' parameter in invert to 0.5 (or higher if you think there are more than 50% bad channels in your selection).

[saracmbr]

Hi Jamie,

I used slop=1 and it worked. It is 30% of the data flagged. Is this a problem I use slop=1 instead of 0.5?

Many thanks,
Sara

[ste616]

Hi Sara,

I wouldn't use slop=1 myself, since this would make invert "use" visibilities that are entirely flagged as bad. In itself, this shouldn't make your image worse, but it could potentially slow down the imaging process.