HSC flagged data for cross-matching
------------------------------------

The required attributes and flagging is described in hsc.pdf.

The chunks directory contains the resultant flagged skinny files for inputting
into cross-matching.

These files were tarred up and copied down to CSD3. 

The initial run didn't have magnitude errors so the code was altered and
re-run to include these.

CSD3 location.

/home/ir-read3/rds/rds-iris-ip005/mar/hsc/hscChunksWithErrors.tar


Columns are:

'object_id','ra','dec','sig_avg','mag0','mag1','mag2','mag3','mag4','bestIndex'
,'magErr0','magErr1','magErr2','magErr3','magErr4','bestFlag','pf0','pf1','pf2'
,'pf3','pf4','chunkId','isOverlap'

where:
ra,dec,sig_avg are from one of  [g/r/i/z/y]_sdsscentroid_[ra/dec]
   depending on the best band from astrometric flagging
mag0-mag4 are the mags in filters order grizy
bestIndex is the umm best index
magErr0-magErr4 the errors on the mags grizy
best flag is the best flag
pf0-pf4 are the individual band photometric flags
chunkID
isOverlap -  0 if core region of chunk, 1 if in overlap

Initially there were some 371651258 sources downloaded from HSC using the
summary table as the base and joining with various other tables to pull out
the required attributes.

Some 333544291 sources made it through the flagging (90%).

Magnitudes are all *_model_mags as no sources were found with |b| <=20.

Notes:

One chunk, 2040, has no overlap sources cos it's an isolated small area of the
survey, AEGIS, entirely covered by the chunk core. 

HSC is comprised of 4 separate regions (including AEGIS). Not sure if that
effects things when it comes to cross-matching or just comes out in the wash?

In photFlagging, sources with magerrs > 0.5 are flagged as 1 but still kept.
Some sources have very large magerrs (> 10 mags), so these are kept,
though the few I've checked are consistent mags with the other bands.