Transition away from WSD_Parks boundary source

[chadlawlis]

• [x] Identify boundaries currently pulling from WSD_Parks equivalent to the latest Lands release and transition to Lands where possible.
  • Unfortunately no boundaries met this criteria...
• Manually compare remaining WSD_Parks to latest Lands release and transition where possible:
  • [x] AKR
  • [x] IMR
  • [ ] MWR
  • [x] NCR
  • [ ] NER
  • [ ] PWR
  • [x] SER

[chadlawlis]

Here is the current status of WSD_Parks presence, by unit_region, ordered by the total number of units pulling from WSD_Parks (to give us a prioritized list for tackling the transition):

unit_region	wsd_parks_count	unit_region_total	wsd_parks_pct
imr	76	87	87
ser	56	68	82
pwr	48	66	73
mwr	46	62	74
ner	44	87	51
ncr	25	39	64
akr	16	17	94

Note: we recently received 32 ncr unit boundaries (https://github.com/nationalparkservice/places-data/issues/347), primarily pulling from Lands and/or a custom park provided source. 22 of the unit codes provided match up with what we have in parks, 18 of which are currently pulling from WSD_Parks. So, this will bring the ncr number down substantially, if not to zero pulling from WSD_Parks.

[chadlawlis]

All 25 NCR units currently pulling from WSD_Parks will be transitioned away from this thanks to https://github.com/nationalparkservice/places-data/issues/347#issuecomment-167001530, so no need to worry about NCR in this manual compare.

[chadlawlis]

See also https://github.com/nationalparkservice/places-data/issues/362/.

[chadlawlis]

See also https://github.com/nationalparkservice/places-data/issues/445.

[chadlawlis]

See also https://github.com/nationalparkservice/places-data/issues/457.

[chadlawlis]

Note: AKR and SER, to date, have indicated that Lands should be the default data source for their unit boundaries (see https://github.com/nationalparkservice/places-data/issues/442#issuecomment-180139897 for AKR, https://github.com/nationalparkservice/places-data/issues/447#issuecomment-182431567 for SER). There are of course exceptions, none for AKR (though two units are missing boundaries from AKR in Lands) but a few from SER (see https://github.com/nationalparkservice/places-data/issues/457#issuecomment-182538113).

PWR, IMR, MWR, and NER have not made any generalizations of defaulting to Lands to date. As per above (https://github.com/nationalparkservice/places-data/issues/348#issuecomment-167002522) NCR will be primarily Lands with some exceptions/customizations.

[chadlawlis]

Current status:

total	wsd_parks	lands	other	null
443	240	132	45	26

via:

select count(*) as total,
(select count(*) from parks where data_source = 'WSD_Parks') as wsd_parks,
(select count(*) from parks where data_source = 'Land Resources Division') as lands,
(select count(*) from parks where data_source != 'WSD_Parks' and data_source != 'Land Resources Division' and data_source is not null) as other,
(select count(*) from parks where data_source is null) as "null"
from parks;

[chadlawlis]

Current status, after NCR update (#347):

total	wsd_parks	lands	other	null
453	201	155	74	23

[chadlawlis]

IMR has been transitioned via #692.

Only OKCI, WACO, and the 5 National Trails remain that do not pull from Lands. We do not have point labels or area geometries for the National Trails, while we have a point and area for WACO from IMR and a point for OKCI from OKCI.

[chadlawlis]

Current status, after IMR update:

total	wsd_parks	lands	other	null
452	126	232	70	24

[chadlawlis]

The data_source field has moved to the parks_poly table in the new places_boundaries_v2 database. To query totals now:

SELECT
    count(*) AS total,
    (select count(*) FROM parks_poly WHERE data_source = 'WSD_Parks') AS wsd_parks,
    (select count(*) FROM parks_poly WHERE data_source = 'Land Resources Division') AS lands,
    (select count(*) FROM parks_poly WHERE data_source != 'WSD_Parks' and data_source != 'Land Resources Division' and data_source is not null) AS other,
    (select count(*) FROM parks_poly WHERE data_source is null) AS "null"
FROM
    parks_poly;