| A catalog of all the objects from Spectro, Best, or Target that "smell" like a QSO|
NOTE: This is a derived science table that may not be loaded for
every data release. It is scheduled to be reloaded for DR5.
| This table is derived from the SpecObj, Best.PhotoObj, and Target.PhotoObj tables. Any SpecObj with SpecClass in QSO or HiZ_QSO is recorded. Any PhotoObj in Best or Target with any of the following primary Target Flags set is recorded: QSO_HIZ, QSO_CAP, QSO_SKIRT, QSO_FIRST_CAP, QSO_FIRST_SKIRT. This list is first 'culled' to discard objects that have multiple observations. The 'primary' (photoObj) or sciencePrimary (SpecObj) is retained. If no primary is available, the highest ObjID of the set wins. Then the three lists are merged into a single list, matching by (ra,dec) with a 1 arcsecond radius. That is, if a photo and spectro object have centers that are within 1 arcsecond, then they are considered to be observations of the same object. This step produces 72319 items on the SDSS DR1 dataset with the following population statistics: |
| Now a sequence of spatial lookups tries to identify the 20,225 best objects that are missing from the first row. Those objects were probably observed, but were not flagged as QSO candidates. After a fair amount of heuristics, most of these corresponding entries are identified. We end up with counts that look more like:|
| The first two rows are great: SpecObjs with matching Best and Target objects, or a matching best and target pair that are likely a QSO but have not yet observed with the spectrograph. The remaining rows are a reminder that this is real data. For example the last row represents a spectrogram with unknown ra,dec. The third and second to last rows represent processing holes in the Best database output (target areas not included in Best.) All this information is recorded in the QsoCatalogAll table along with:|
|Index Type||Key or Field List|
|primary key||zone, ra, SpecObjID, bestObjID, targetObjID|