Create DR13 APOGEE-RC sample

[jobovy]

Basic commands:

python make_rcsample.py -o ~/tmp/apogee-rc-DR13-addl.fits --rmdups --nostat --nopm --addl-logg-cut

python check_rc_against_apokasc.py ~/tmp/apogee-rc-DR13-addl.fits

Need to do the following:

• [x] Check contamination compared to DR12
• [x] Cross-match against DR12 to see how different they are (same input data!)
• [x] Decide on whether to apply addl-logg-cut
• [x] Cross-match against Tycho
• [x] Make plots from RC paper to see how they've changed
• [x] Work on datamodel
• [x] Get on the SAS
• [x] Check and update documentation

[jobovy]

Using checks wrt APOKASC:

DR13 catalog (no addl-logg-cut)

• 19,291 stars
• 7% contamination --> 5% adjusted
• 44% completeness

DR13 catalog (w addl-logg-cut)

• 18,417 stars
• 5% contamination --> 3.5% adjusted
• 43% completeness

DR12 catalog (w addl-logg-cut)

• 19,937 stars
• 14% contamination (!!) --> 10% adjusted
• 31% completeness

Conclusion: Don't use ADDL_LOGG_CUT by default

[jobovy]

Comparison to DR12

• 14,288 stars in common: ~5,000 different --> even at ~15% total contamination, seems high.
• Metallicity has shifted down by ~0.1 dex, that looks correct based on previous calibration offset
• [alpha/Fe] has shifted down by -0.03 dex, exactly what M67's offset was before!
• Distance distribution almost exactly the same (off in the median by 2 pc!); l and b also the same