R/forecast_infections.R
forecast_infections.Rd
Provides optional tools for forecasting cases and Rt estimates using the timeseries methods
(via the
EpiSoon
package). It requires the Episoon
package. Installation instructions for the EpiSoon package are
available here.
forecast_infections(
infections,
rts,
gt_mean,
gt_sd,
gt_max = 30,
ensemble_type = "mean",
forecast_model,
CrIs = c(0.2, 0.5, 0.9),
horizon = 14,
samples = 1000
)
A data frame of cases by date of infection containing the following variables: date, mean, sd
A data frame of Rt estimates by date of infection containing the following variables: date, mean, sd
Numeric, the mean of the gamma distributed generation time.
Numeric, the standard deviation of the gamma distributed generation time.
Numeric, the maximum allowed value of the gamma distributed generation time.
Character string indicating the type of ensemble to use. By default this is an unweighted ensemble ("mean") with no other types currently supported.
An uninitialised forecast model function to be passed to EpiSoon::forecast_rt
. Used
for forecasting future Rt and case co An example of the required structure is: function(ss, y){bsts::AddSemilocalLinearTrend(ss, y = y)}
.
Numeric vector of credible intervals to calculate.
Numeric, defaults to 14. The horizon over which to forecast Rts and cases.
Numeric, the number of forecast samples to take.
A list of data.tables
. The first entry ("samples") contains raw forecast samples and the second entry ("summarised") contains
summarised forecasts.
# \donttest{
if (requireNamespace("EpiSoon")) {
if (requireNamespace("forecastHybrid")) {
# example case data
reported_cases <- example_confirmed[1:40]
generation_time <- get_generation_time(disease = "SARS-CoV-2", source = "ganyani")
incubation_period <- get_incubation_period(disease = "SARS-CoV-2", source = "lauer")
reporting_delay <- estimate_delay(rlnorm(100, log(6), 1), max_value = 15)
# estimate Rt and infections from data
out <- estimate_infections(reported_cases,
generation_time = generation_time,
delays = delay_opts(incubation_period, reporting_delay),
rt = rt_opts(prior = list(mean = 2, sd = 0.1))
)
# forecast Rt and infections from estimates
forecast <- forecast_infections(
infections = out$summarised[variable == "infections"],
rts = out$summarised[variable == "R"],
gt_mean = out$summarised[variable == "gt_mean"]$mean,
gt_sd = out$summarised[variable == "gt_sd"]$mean,
gt_max = 30,
forecast_model = function(y, ...) {
EpiSoon::forecastHybrid_model(
y = y[max(1, length(y) - 21):length(y)],
model_params = list(models = "aefz", weights = "equal"),
forecast_params = list(PI.combination = "mean"), ...
)
},
horizon = 14,
samples = 1000
)
forecast$summarised
}
}
#> Loading required namespace: EpiSoon
# }