inst/dev/design_s3.md

In epiforecasts/EpiNow2: Estimate Real-Time Case Counts and Time-Varying Epidemiological Parameters

S3 return types for model types

All three models to return an object (epinowfit or the like) inheriting from list and containing

• fit (the stan object)
• args (the arguments that have been passed to stan)
• observations (the timed data frame that has been passed as observations)

with an appropriate print() method that shows some appropriate summary of what the object contains (no results).

Mapping of existing return elements to accessor functions

estimate_infections

• samples: get_samples() (or extract_samples()) built on the existing extract_parameter_samples(); this could also take a parameter name and look up the ID, or take a delay distribution name and look up the ID;
• summarised: summary()
• fit: get_stan_fit()
• args: not needed (only internally to identify parameter ids)?
• observations: not needed (only internally by summary() and get_samples()

estimate_secondary

• predictions: get_predictions? or predict()? This could be used for all three models to get posterior predictions, which would help with a few other things (plotting, scoringutils integration...)
• posterior: same as samples in estimate_infections
• data: same as observations in estimate_infections
• fit: same as fit in estimate_infections

estimate_truncation

• dist this will have to be accessed using the summary or get_samples() - we could add an extra accessor function to get a dist_fit() using mean and sd of samples if we think this is a good idea
• obs: same as observations in estimate_infections
• last_obs: this is really no needed as contained in obs
• cmf: not needed as can be obtained from dist (and is already included in plotting <dist_spec>)
• data: same as observations in estimate_infections
• fit: same as fit in estimate_infections

epiforecasts/EpiNow2 documentation built on June 9, 2025, 3:51 p.m.