Handling missing variables from codebook that are present in data and crosswalk files

[maqifrnswa]

First off, thank you! This project is so helpful!

I came across a few variables that are missing from PSIDcodebook.xml but are in the data and psid.xls. I know you don't maintain any of those files, but it does cause makePSID to error out. Should there be a fallback in those cases? Maybe if it is present in the .xls, but not in the xml, it still processes everything but leaves the _label column blank? I think that's the only consequence of not having the codebook.

Example variables that are missing in the xml files but are in the xls file:

• V13500 (spouse, race mention 1). Workaround: It's identical to V13501, so I just edited the xml file to copy that and changed references to race 2 to race 1
• ER38263A1, ER44236A1, ER39360A1, ER45333A1 (head/reference and spouse "other chronic conditions"). Workaround: These are the same as ER49574 or ER50692.

[aaowens]

It is a good idea to not error and leave the _label blank. I'll look at it in a week or so.