geocode geo_loc_names from MongoDB prod and from SubmissionPortal ("prod"?)

[turbomam]

Google geolocation price: https://developers.google.com/maps/documentation/geolocation/usage-and-billing

0.005 USD per each (5.00 USD per 1000) @ up to 100,000 per month

As of today, there's about 350 unique normalized addresses in MongoDB prod and SubmissionPortal prod

So about $1.75 to process the whole set

up-to-date geocodings are saved as:

https://github.com/microbiomedata/submission-schema/blob/issue-151-geocoding/geocoded_normalized_geo_loc_names.yaml

---

• we are asserting 'sample_collection_site' values on the NEON biosamples, like 'ONAQ'
• would really be better to use one of these formalisms:
  • https://microbiomedata.github.io/nmdc-schema/sample_collection_site/
  • https://microbiomedata.github.io/nmdc-schema/Site/
  • https://microbiomedata.github.io/nmdc-schema/FieldResearchSite/
    • Biosample collected_from range FieldResearchSite
  • https://microbiomedata.github.io/nmdc-schema/CollectingBiosamplesFromSite/
• want slot name included in DH help screens
• want structured pattern and string serialization included
• we may want the number of digits in latitudes and longitudes influence our trust of them
  • lat lon decimal places significance : https://support.garmin.com/en-US/?faq=hRMBoCTy5a7HqVkxukhHd8
• google results incude bounding box. really large bounding boxes (like when geocoding a US state), may indicate low-precision geo_loc_names

[turbomam]

add reverse geocoding?

[turbomam]

https://maps.googleapis.com/maps/api/geocode/json?latlng=40.714224,-73.961452&key=YOUR_API_KEY

[turbomam]

https://maps.googleapis.com/maps/api/elevation/json?locations=39.7391536%2C-104.9847034&key=YOUR_API_KEY

[turbomam]

why does one endpoint recommend using a bare , and he other recommends encoding as '%2C' ?

https://www.asciitable.com/

, does work in both cases