Provider site reference table:

[Joeyloganpython]

There is a table that relates many sites to many providers and vice vera: Is there some sort of data for this?

[jkdru]

Sites and providers each have several of their own datasets.

As with waivers, I think addresses (and sometimes also phone) are the key link between providers and sites. The relationship is probably better handled with views than a table per se.

Could we try loading sites tables together when we meet in a few days?

[jkdru]

Related to issue #10

By Sept 3 AM (EDT), will provide separate spreadsheet tabs for at least two _listings ("rec") tables:

SITES
-- samhsa_site_listings (compare Aug 24 ("Behavioral_Health...") fresh download with Apr 29 compilation)
-- dbhids_site_listings
-- samhsa_site_codes

PROVIDERS
-- samhsa_provider_listings
-- npi_listings (just combine main lode with 15-25 recent additions)
-- pals_listings (change colname to moreinfo_seePALS, retain all recs per accurate person, use 0 for empty top_one rows)
-- xwaivers (use provider ID as PK; add col for past_dea_num and missing cert_state, cert_date cols)

[jkdru]

@Joeyloganpython, should the new data loads for provider _listings incorporate the P-----## IDs already generated in the test loads or let postgreSQL generate them afresh?

-- sites table itself is needed to generate site IDs, but has similar chicken/egg issue as providers (rough model, don't use as-is)
-- sites come from multiple rec sources (whereas providers currently come only from samhsa_provider_listings)

[jkdru]

@Joeyloganpython, @atzbencam: Here are the sheets for nearly load-ready datasets -- one tab per table. Could we look them over tomorrow? Should the new data loads for provider _listings incorporate the P-----## IDs already generated in the test loads or let postgreSQL generate them afresh?

[jkdru]

Closing this issue to declutter -- will echo keeper points in updated issue for DB 2.0 (Jan/Feb 2021)