search

cmu-delphi / covidcast-indicators

Back end for producing indicators and loading them into the COVIDcast API.
https://cmu-delphi.github.io/delphi-epidata/api/covidcast.html
MIT License
12 stars 17 forks source link

Nhsn indicator #2080

Open aysim319 opened 3 days ago

aysim319 commented 3 days ago

Description

Adding new indicator NHSN

Associated Issue(s)

nmdefries commented 1 day ago

Waiting on decisions from tooling team on what smoothing and geo aggregations we should do, and if we should include any other metrics (currently generates total flu and covid hospital admissions. Source data also includes metrics by age group, but we aren't ingesting those).

RE geo aggregation, the previous hhs indicator only pulled state and nation data, and didn't do any geo aggregation. This indicator does the same thing.

@dshemetov @brookslogan @dsweber2 for input.

nmdefries commented 1 day ago

Due to production issues, will aim to release this early next week.

aysim319 commented 1 day ago

context from david:

oh, also, dunno if you saw the post from logan, they're doing this really dumb thing of putting the preliminary data in a separate source https://data.cdc.gov/Public-Health-Surveillance/Weekly-Hospital-Respiratory-Data-HRD-Metrics-by-Ju/mpgq-jmmr/about_data we should probably include those as separate versions in the same indicator oh er, when I said versions, I meant versions (like the release date) rather than different signals. Definitely not the best wording choice.

Currently have it such that they're seperate signals similar to nssp secondary source to differienate them as a place holder

dsweber2 commented 23 hours ago

Working through this. I tried a couple of smoothings (3 & 4 weeks, both trailing and centered) to see if that would help tame some of the more extreme data issues, but it didn't really help. I had a use-case where hhs_regions would've been handy, so that's probably worth adding eventually (probably more important to get this live first though).