Question: RA_max / DEC_max limited to pixel grid?

[Jordatious]

Hi all, I note from the docs that when gaussian components are combined, "the maximum of the source" is stored (presumably as RA_max and DEC_max), as well as the centroid. Does the maximum refer to the (interpolated) position of the brightest gaussian component, or simply the position of the brightest pixel?

[darafferty]

I had to check the code, but I believe the maximum generally corresponds to the position of a Gaussian fit to the region around the maximum of the source (see https://github.com/lofar-astron/PyBDSF/blob/master/bdsf/gaul2srl.py#L356 where the fit is done). So it should not be limited to the pixel grid.

[Jordatious]

Thanks for the info @darafferty. My use case here is identifying the central (brightest) position of bright outliers. Currently, I'm taking the position of the brightest Gaussian component within a small cutout as my position of interest. But it sounds like this is the same value as RA_max and DEC_max from the source catalogue. But perhaps the source grouping parameters affect this.

My current workflow is fine, but I wanted to know if the maximum was fixed to the pixel grid. You've answered that, so feel free to close the ticket, but any other clarity you can provide would be helpful!