Skip to contents

Summarize parameters of interest

Source: R/sewer_summary.R
summarize_fit.Rd

This function summarizes important parameters of interest from a fitted EpiSewer model.

Usage

summarize_fit(fit, data, .metainfo, intervals = c(0.5, 0.95), ndraws = 50)

Arguments

fit

The fitted EpiSewer model.

data

The model data that the EpiSewer model was fitted with.

.metainfo

Meta information about the model data.

intervals

The credible intervals (CrIs) that should be calculated. By default, these are the 50% and 95% CrIs.

ndraws

Number of exemplary posterior samples that should be extracted. Note that the summaries always use all available samples.

Value

A list with the following summaries (each in a data.frame):

  • R (posterior summary)

  • R_samples (exemplary samples)

  • expected_infections (posterior summary)

  • infections (posterior summary)

  • infections_samples (exemplary samples)

  • expected_load (posterior summary)

  • expected_concentration (posterior summary)

  • concentration (posterior summary)

  • sample_effects (posterior summary)

Details

Normally, this function is automatically called by EpiSewer after model fitting, but it may also be run manually, for example to update the summary intervals or number of exemplary samples. Note that this is only possible if the result contains the full fitted model object (see set_results_opts()).

The summaries for infections and R include a column seeding, which indicates whether the corresponding date was still in the seeding phase or not. The seeding phase is here defined as 2xG days long, where G is the maximum generation time. This is in contrast to the model, where it is only G days long. The reason for this decision is that after G days, new infections are still based on the seeded infections and not strongly informed by the data. By using the extended seeding criterion, we ensure that only sufficiently data-driven estimates are shown in the result plots.