CPCS matching radius fix

[kkotysz]

Change fixed matching radius == 2 arcsec to more flexible parameter tied to sigma of astrometric solution. Examples of well reduced datums: one matched and one not matched below.

• https://bh-tom2.astrolabs.pl/dataproducts/data/263446/ -> matched
• https://bh-tom2.astrolabs.pl/dataproducts/data/263360/ -> not matched

TOI is conveniently marked in .dat file with 999999 object number.

[wyrzykow]

The idea is that we will use the actual matching radius based on the Ra-Dec cleaning we do at the very beginning of CPCS. We will then use sqrt(deltaRA^2+deltaDec^2) x 1.5 as the matching radius when identifying a target in the get_results method. Simple fix, but requires attention in coding.

[wyrzykow]

Coded, added to cpcs-test. Matching radius is found from Ra-Dec match and is used for target identification. I use 2x sigma_ra_dec, but we can then see if that is too large.

Match_dist is stored in the db by the uploader, but then its value is updated after found by ra-dec. We store match_dist in arcseconds. Note the db contains a typo in the column name: match_distans, but it has to stay.

Still need to test it for Gaia19eyy.

[wyrzykow]

Tested on Gaia19eyy. Computed the sigma of radec to 2.6: from cpcs log: -matchingdist used: 5.220130901027017 Target search: minimum distance found of 2.1321952094990215 arcsec

Works! Installed on cpcs-test. Can be moved to prod.