10 add hosp admissions only model functionality - Githubissues

CDCgov / ww-inference-model

An in-development R package and a Bayesian hierarchical model jointly fitting multiple "local" wastewater data streams and "global" case count data to produce nowcasts and forecasts of both observations

https://cdcgov.github.io/ww-inference-model/

Apache License 2.0

18 stars 2 forks source link

10 add hosp admissions only model functionality #24

Closed kaitejohnson closed 4 months ago

kaitejohnson commented 4 months ago

This PR closes #10 and also has a bit of scope creep to fix the rendering of the vignette/some bugs in the figure generating function. Changes include:

adding a component to the vignette that fits the model to only hospital admissions and plots the resulting hospital admissions forecasts
adjusted the generate simulated data function and resulting package data to include a larger population size, bc the low counts of admissions were causing diagnostic issues
modified get_model_spec() and get_stan_data() to take in include_ww as an argument, to specify whether or not to fit the model to wastewater data
adjusted the way the figures were rendering in the vignette (mirroring epinowcast vignette to point to a figure folder which contains figures that render the width of the screen

kaitejohnson commented 4 months ago

Note this doesn't do what I mentioned in the MSR community meeting today -- it doesn't add the functionality to not have wastewater data. I will do this in a separate issue... I think by making an empty dataset

kaitejohnson commented 4 months ago

@dylanhmorris not a huge rush but this is ready for review and is relatively small (tweaks to vignette + the easy version of hospital admissions only functionality)

kaitejohnson commented 4 months ago

Although I am realizing now that I set up the github branch protections to require status checks to pass, which includes deploy of pkgdown but I think deploy wouldn't happen until the PR gets merged to main so maybe I should remove that one

dylanhmorris commented 4 months ago

Although I am realizing now that I set up the github branch protections to require status checks to pass, which includes deploy of pkgdown but I think deploy wouldn't happen until the PR gets merged to main so maybe I should remove that one

Correct you should require a successful build but not a successful deploy

kaitejohnson commented 4 months ago

Although I am realizing now that I set up the github branch protections to require status checks to pass, which includes deploy of pkgdown but I think deploy wouldn't happen until the PR gets merged to main so maybe I should remove that one

Correct you should require a successful build but not a successful deploy

@dylanhmorris Would you mind taking a look at this and the ruleset? I tried to specify just pkgdown / build (pull_request) in the require status checks before passing but for some reason it still says expected, status waiting to be reported despite the successful build above