Still a work in progress;
- financial tables f2 / f3 have some issues ; need to do some cleanup on the variables list.
- one variable (CNFNO4) for some reason not being found (exists in table being read).
- The original notebook (make a master merge table of everything first) fails (hangs the kernel) after about 11 files merged. (I find this peculiar?) Appears to be a memory issue (thogh the fiels are not huge). THe findmerge variant (not working fully yet as of 12/20) searches for variables of interest FIRST in each table one at a time as they are read, then with the results, attempts to build a merged dataframe with only the variables we want.
- Many of the tables have two versions
- table.csv
- Initial submission of responses to IPEDS
- Happens at the end of the year; available the next year (usually)
- table_rv.csv
- This appears to be the version that IPEDS reviews a year after the first submission
- We should use this version if it is available (won't be available for the most recent year submission)
- Not every table gets changed/reviewed in which case there is only the initial table
- table.csv