DB_QSO_S82.dat – file holding all qsos in Stripe 82 (DR7 release)
pbh_gibsontab1ANDallenANDdr7q.dr7bh – dr7 RA/DEC for the GIBSON BAL quasar list …

need to cross reference.

looks like IDL is not happy with matching up RAs and DECs (Rounding Issues) Want to make sure I’m doing this right. Going to set a tolerance of roughly 1 arcsec and look for matched in IDL.

Using a tolerance of 0.0003 degrees for BOTH RA and DEC: This is roughly \pm 1 arcsec (one arcsec is 0.00027777~). I forced both values in the list to be FLOATs in the comparison, which does force some rounding. In matching the lists in this way, I find that 786 of the 9275 objects in S82 are BAL quasars. That’s 8.5% – which is a little low. My understanding is that BAL quasars are around 15% of all quasars. With a sample as large as almost 10 000, I would expect that percentage to be upheld.

Casjobs has a way of selecting matching values in columns, so as an excercise in both learning CasJobs, and confirming my BAL list in S82, I’ll check that out as well.

SUCCESS: CasJobs and IDL agree that 786 quasars from S82 list by MacLeod are BALs (as per matching to Gibson list). Is CasJobs based on IDL?

Two things to check – are both lists the SAME, and are there repeats? Does Gibson list have multiple iterations of the same BAL in the list? I suspect that might be true. Will inflate the percentage I get.

1) repeats
there should be no repeats in the S82 data from MacLeod. So I’ll check that first.
IDL code shows no matches. Meaning Each RA and DEC in the S82 list is unique (knew that already). now do gibson list.

Gibson has repeats. Which I also knew, looking back on README files on Pat’s website. 5511 entires on 5037 objects. Presumably that means there 474 repeated lines in the Gibson list. Using IDL I find the list has 625 repeated objects. Problem.

But i think this can occur if there are more than two repeats of the same object. That would inflate the ‘repeated’ objects, without actually having more. So to account for the repeated objects, I have to not only check for matches, but then check to see if a match has already been made, and if so, forget that one.

2)check lists from CasJobs and IDL to make sure they’re all the same objects being matched.
Still working on this, but it appears at the moment that the two lists (from IDL and from CasJobs) are different in nature. Could be a programming effect. But it’s telling that both methods find the same number…isn’t it?