Present: Lauren MacArthur, Unknown User (tmorton), Sophie Reed, Jim Bosch, Paul Price.
- Tim has created the parquet tables for all the PDR1 data.
- We looked at the ‘principle color’ stellar locus measurements for one of the wide fields, and saw spatial variation and correlated badness between the various colors.
- Jim suggests that the next most useful feature to have available for more drill-down with my QA toolkit would be to have visit/CCD outlines drawn on the sky plots, so we can see whether boundaries of spatially-varying badness correspond to particular visits. I brainstormed a few different ways we might be able to do that interactively without allowing the plots to get too busy.
- We discussed matching a bit; Sophie has sent Tim some code that implements a kd-tree based matching, and I plan to implemented a
MatchedCatalog
object that will accomplish position-based matching between two of my customCatalog
objects. This will enable me to make difference plots of quantities and to calculate metrics relating to repeatability.