Open tjgalvin opened 2 months ago
Happy to hear suggestions otherwise.
I suppose the major tradeoffs are:
Thinking aloud here. I am already having to handle catalogues at some level in order to generate the validation figures and tables. At the moment these are not really tracked by either the flint git, nor are they downloaded by flint on first start up.
Trying to package something does help with testing and crafting something that is reproducible. Not to say this can not be done with astroquery.
At the moment it would be annoying to move away from the referenced set of sources, as there is a column that includes the size of the FoV to image of candidate sources to peel. Useful for the larger resolved things out there.
In a recent flint update the functionality to peel sources using @sunmish's potatopeel software was added. This iterates over a list of sources, evaluates whether they should be peeled and/or subtracted, and will either peel or subtract them.
At the moment the list of candidate sources to peel are based on a package set of sources derived by @sunmish (complete with appropriate aperture sizes) that was constructed throughout the processing of racs-mid and racs-high.
Perhaps this should be expected to use astroquery to get a list of radio sources within a few degrees of a beam center, or package a subset of racs-mid made up of bright sources, that are then used as the basis for peeling.
At the moment the two criteria that are used to evaluate whether a source is provided to potatopeel for examination is:
There are currently no constrains in
flint
to come up with the candidate set of sources. Potatopeel will internally examine the apparent brightness of the source and make its own determination using some very basicflint
provided thresholds.