trajectories_to_binned_distributions: Convert a matrix of sampled trajectories to binned...

In reichlab/cdcForecastUtils: Utility functions for CDC forecasts

Description

Convert a matrix of sampled trajectories to binned distributions for short-term and/or seasonal targets.

Usage

trajectories_to_binned_distributions(
  trajectories,
  targets,
  h_max,
  bins,
  baseline,
  season_start_ew,
  season_end_ew,
  cdc_report_ew
)

Arguments

trajectories	matrix of simulated trajectories. Each row is one simulated trajectory, each column is one time point. Must contain all time points needed to calculate seasonal and short-term targets
h_max	largest horizon for short-term targets
targets:	character vector specifying targets to compute. May include: "wk ahead", "Below baseline for 3 weeks", "First week below baseline", "Peak height", "Peak week"
bins:	vector of start and end points for incidence targets. For example: c(seq(from = 0.0, to = 25.0, by = 0.1), 100.0)
baseline:	baseline for this unit, if applicable; required if "onset timing", "Below baseline for 3 weeks", or "First week below baseline" are among the targets to forecast
season_start_ew:	character specification of season start epidemic week, formatted as "2019-EW40"
season_end_ew:	character specification of season end epidemic week, formatted as "2020-EW20"
cdc_report_ew:	character specification of epidemic week corresponding to most recent cdc report, formatted as "2019-EW40"

Value

data frame with columns: target: with values coming from targets type: populated with "Bin" bin: name of bin for categorical targets, lower endpoint of bin for numeric targets value: proportion of trajectories falling in bin

reichlab/cdcForecastUtils documentation built on May 6, 2020, 10:43 a.m.