DE notes on 22Q2

[AmandaDoyle]

• Job_inactive field needs to be normalized (i.e. there are inactive and Inactive values)
• 2 records WHERE classaprop-classainit <> classanet (Jobs: 320755483 and 520328422)
• datepermit and datecomplt are in a different date format than the other date fields

[td928]

took a look at the job 320755483 and 520328422 and my intuition was that it might be due to manual corrections. Turns out they both are records got corrected and proposed and initial values actually align with the post-correction value but for some reason the classa_net is calculated incorrectly and it can be traced back to the units_devdb table. Need to investigate further within _units.sql to find exact reason of breaking.

UPDATE: Okay solved 520328422. The current logic dictates that when it is New Building job type, the classa_prop will be automatically be the final value classa_net. That's why the 1 unit is not taken out of the 20 units.

Also similarly for 320755483, it is demolition type so that the classa_net is -1 * classa_init = 0

[mbh329]

Didn't notice anything that stood out to me but interesting to see that far rockaway had the largest increase in classa units

[td928]

version needs to be changed to 22Q2