715 records in the PUF have n65>n21

[MaxGhenis]

Defining n65 as (age_head>64) + (age_spouse>64) + elderly_dependents, 715 records have n65>n21. All 715 have age_head>64, and they otherwise fall into 3 categories (notebook):

1. 700 records have n21=1 and age_spouse>64, e.g. RECID 415.
2. 13 records have n21=2, age_spouse>64, and elderly_dependents=1, e.g. RECID 6869.
3. 2 records have n21=1, and elderly_dependents=1, e.g. RECID 13275.

[MaxGhenis]

I think a way to avoid this is to impute n2164 separately from the CPS, then add it to a calculated n65 to produce n21.

[andersonfrailey]

I think that's a good idea, @MaxGhenis. I'm working on refactoring the statistical matching process now. I can work that in as I go.

[MaxGhenis]

Was this addressed in recent updates?

[andersonfrailey]

@MaxGhenis, yep this has been resolved with the latest updates. Closing.