Test complete filling for collected condo chars

[wrridgeway]

See https://github.com/ccao-data/data-architecture/issues/309.

This seems like a better solution for the present then back/forward filling these chars since we only have one observation per pin. Once this data gets moved into iasworld we can start filling. Until then, this is probably more robust.

[wrridgeway]

with old as (
    select
        year, count(*) as old
    from default.vw_pin_condo_char
    group by year
),

new as (
    select
        year, count(*) as new
    from "z_ci_309-explore-collected-characteristics-filling-for-condos_default".vw_pin_condo_char
    group by year
)

select
    old.*,
    new.new
from old
left join new on old.year = new.year
order by year desc

year	old	new
2024	457640	457640
2023	457640	457640
2022	459126	459126
2021	458980	458980
2020	458449	458449
2019	458673	458673
2018	459473	459473
2017	459275	459275
2016	458847	458847
2015	458355	458355
2014	458049	458049
2013	457812	457812
2012	458194	458194
2011	457166	457166
2010	454939	454939
2009	445790	445790
2008	429185	429185
2007	402508	402508
2006	370734	370734
2005	342996	342996
2004	321082	321082
2003	300884	300884
2002	283450	283450
2001	266830	266830
2000	250302	250302
1999	235795	235795

[wrridgeway]

Sure. My main motivation here was that it's more likely than not that this data will make its way into iasworld before we have multiple years worth of collected data for the same PINs (if this is even something that happens) in the current format and will need a complete refactor regardless.