mgraber
commented
3 years ago Top coding threshold for 2018 & 2019
- mdage: 85 (yes, same as lower limit of top bin)
- mdhhinc: 200000 (yes, same as lower limit of top bin)
- mdfaminc: 200000 (yes, same as lower limit of top bin)
- mdnfinc: 200000 (yes, same as lower limit of top bin)
- mdewrk: 100000 (yes, same as lower limit of top bin)
- mdemftwrk: 100000 (yes, same as lower limit of top bin)
- mdefftwrk: 100000 (yes, same as lower limit of top bin)
- mdrms: 9000 (not top coded)
- mdvl: 2000000 (yes, same as lower limit of top bin)
- mdgr: 3500 (yes, same as lower limit of top bin)
Top coding threshold for 2010 (non-inflated)
https://github.com/NYCPlanning/db-factfinder/issues/162#issue-947955600
- mdage: 85 (yes, same as lower limit of top bin)
- mdhhinc: 200000 (yes, same as lower limit of top bin)
- mdfaminc: 200000 (yes, same as lower limit of top bin)
- mdnfinc: 200000 (yes, same as lower limit of top bin)
- mdewrk: 100000 (yes, same as lower limit of top bin)
- mdemftwrk: 100000 (yes, same as lower limit of top bin)
- mdefftwrk: 100000 (yes, same as lower limit of top bin)
- mdrms: 9000 (not top coded)
- mdvl: 1000000 (yes, same as lower limit of top bin)
- mdgr: 2000 (yes, same as lower limit of top bin)
Based on labs thresholds: https://github.com/NYCPlanning/labs-factfinder-api/blob/f8033933664f29d07b1ee9a04792b6f31e0241a3/special-calculations/data/top-bottom-codings.js
This is a modification of the following lines, which currently bottom code all medians: https://github.com/NYCPlanning/db-factfinder/blob/66094c6178ab42392552096539c1ba9065450953/factfinder/median.py#L36-L38
Top coding occurs here, again for all medians: https://github.com/NYCPlanning/db-factfinder/blob/66094c6178ab42392552096539c1ba9065450953/factfinder/median.py#L42-L44