Taisiya Kopytova (ASU) arrived in New York today for two days to work on the detailed chemical abundances of giant stars in APOGEE that host substellar and low-mass stellar companions. We have cast this problem as a very simple comparison between the companion-hosting population and a control sample. The control sample is carefully selected to be identical in stellar parameters. It looks like she has a result, and the result is interesting. It isn't a large effect, however. We discussed how to show it, convincingly, and how to describe it, accurately.
Lauren Anderson (Flatiron) and I continued to debate how to put spectroscopic labels onto Gaia TGAS stars without spectroscopy, and how to (before that) de-noise the labels themselves, which we are getting from the RAVE-on catalog of Andy Casey et al. We are a bit confused about how principled to be in the latter de-noising. It seems crazy to proceed without doing it, but building a full hierarchical model seems like overkill when the main point is just that stars on the main-sequence must have large surface gravities!