README.md

In Bisaloo/NpiEurope: What the Package Does (One Line, Title Case)

NpiEurope

This code is used to produce the results from the paper "Quantifying the efficiency of non-pharmaceutical interventions against SARS-COV-2 transmission in Europe", by Garchitorena et al (https://www.medrxiv.org/content/10.1101/2020.08.17.20174821v1)

Installation

The development version can be installed from GitHub with:

# install.packages("remotes")
remotes::install_github("Bisaloo/NpiEurope")

Example

library(NpiEurope)

Let’s first load the data we’ll need for this example. In this example, we focus on the case of Poland.

country <- "Poland"

The functions load_country_data(), load_contact_data() and load_age_date() are helpers for the handful of countries that have been bundled with this package but you can provide your own dataframe for country_data, contact_data and age_data. The load_country_data() can also take an optional dataset_path arguments where you can specify the path for the ECDC dataset in the case when the version bundled with this package gets out of date.

country_data <- load_country_data(country)
#> Parsed with column specification:
#> cols(
#>   .default = col_double(),
#>   un = col_character(),
#>   Country = col_character(),
#>   Date = col_date(format = "")
#> )
#> See spec(...) for full column specifications.
country_data <- estimate_asympto(country_data)
plot_asympto(country_data)

contact_data <- load_contact_data(country)
#> New names:
#> * `` -> ...1
#> * `` -> ...2
#> * `` -> ...3
#> * `` -> ...4
#> * `` -> ...5
#> * ...
age_data <- load_age_data(country)

You can estimate epidemiological parameters (proportion of asymptomatic cases propAsympto, transmission rate transmRate and efficiency of the lockdowns vecEff) via the LHS_init() function which takes the country as argument:

res <- LHS_init(country_data, contact_data, age_data, task = "estimate")

The number of particles and number of iterations for the MCMC can be ajusted with the Np and Niter arguments respectively.

You can view the trace of the MCMC with:

plot_MCMC(res)

You can see the 95% credibility interval of each estimated parameter with the function summarise_estimation(). This function prints the credibility intervals:

summarise_estimation(res, country_data)

but also returns a table with these intervals so you can build a nice table from it, e.g., with knitr::kable():

knitr::kable(summarise_estimation(res, country_data))

and then visually compare the simulated trajectories from these parameters with the empirical data. The Npost argument in simulate_trajectory() specifies how many parameter values you want to extract from the posterior distribution of the MCMC to run the simulations:

sims <- simulate_trajectory(res,
                            country_data, contact_data, age_data,
                            Npost = 5)
plot_simulations(sims, country_data)

Bisaloo/NpiEurope documentation built on Dec. 2, 2020, 12:14 p.m.