what are unassigned cases?

[rtwfroody]

• Source: https://liproduction-reportsbucket-bhk8fnhv1s76.s3-us-west-1.amazonaws.com/v1/latest/timeseries-byLocation.json
• Issue: E.g. there's an entry for unassigned cases in Maine. What are they?

  {
    "stateID": "iso2:US-ME",
    "tz": "America/New_York",
    "slug": "unassigned-maine-us",
    "population": 1344212,
    "stateName": "Maine",
    "countryName": "United States",
    "name": "Unassigned cases, Maine, US",
    "level": "state",
    "countryID": "iso1:US",
    "locationID": "iso1:us#iso2:us-me#(unassigned)",

[jingjtang]

Are those unassigned cases back-filled? If they are, where could we find the information about the timing that you add the unassigned info? (I noticed that the unassigned numbers don't exist for the most recent date but exist for the previous dates.)

[jzohrab]

Hi @rtwfroody and @jingjtang - sometimes states report cases etc. but don't assign those to a known county.

e.g., for Maine, per the code (src/shared/sources/us/me/index.js), we're scraping a google sheet (https://docs.google.com/spreadsheets/d/13Rbm5zKKLTFNyLZ2Z9YYHc5v6YpO_erMz1pZwiUtfiQ/gviz/tq?tqx=out:csv&sheet=cases_by_county) referenced from the government site https://www.maine.gov/dhhs/mecdc/infectious-disease/epi/airborne/coronavirus/data.shtml. That google sheet contains the following:

"PATIENT_COUNTY","DATA_REFRESH_DT","DATA_AS_OF_DT","CASES","DEATHS","RECOVERIES","HOSPITALIZATIONS"
"Androscoggin","2020-07-30","2020-07-29","547","6","481","40"
...
"Somerset","2020-07-30","2020-07-29","34","1","31","2"
"Unknown","2020-07-30","2020-07-29","1","0","0","0"
"Waldo","2020-07-30","2020-07-29","62","14","47","5"
...

We're mapping "Unknown" to "Unassigned", a common thing that we do throughout our sources.

@jingjtang - re "Are those unassigned cases back-filled?" , we're pulling those from our source data only. We don't back-fill anything. We do occasionally refresh our "timeseries" data, meaning we re-scrape everything, but that's not really backfilling.

I think this answers your questions, so back to you @rtwfroody to close this. If you have more q's, let me know! Cheers, jz

[rtwfroody]

Thanks for the explanation!