SimulatedForecast: SimulatedForecast

In HopkinsIDD/ForecastFramework: A Basis for Modular Model Creation

Description

This class is a forecast where the data is many simulated trials.

Fields

data

The data used to create the forecast.

forecastMadeTime

When the forecast was created.

forecastTimes

The times the forecast is about.

model

The model used to create the forecast.

nsim

The number of simulations.

sample

Draw a random sample from the possible model predictions. Please see implementation of the data for the properties of the sampling.

Methods

binDist(cutoffs,include.lowest = FALSE,right = TRUE)

Get the distribution of simulations of the data within fixed bins.

Arguments

cutoffs	-	A numeric vector with elements to use as the dividing values for the bins. -Inf, and Inf will be added automatically.
include.lowest	-	logical, indicating if an x[i] equal to the lowest (or highest, for right = FALSE) breaks value should be included.
right	-	logical, indicating if the intervals should be closed on the right (and open on the left) or vice versa.

Value

an ArrayData.

debug(string)

A function for debugging the methods of this class. It calls the browser command. In order for methods to opt into to debugging, they need to implement the following code at the beginning: if(<method_name> %in% private$.debug){browser()}. This method exists, because the debugger is not always intuitive when it comes to debugging R6 methods.

Arguments

string	-	The name(s) of methods to debug as a character vector

initialize(...)

This function should be extended. Create a new instance of this class.

Arguments

...	-	This function should take in any arguments just in case.

mean()

This method extracts the elementwise mean of the forecast. This function will not change the number of rows or columns in the data, but will convert probabilistic estimates into deterministic ones.

Arguments

median()

This method extracts the elementwise median of the forecast. This function will not change the number of rows or columns in the data, but will convert probabilistic estimates into deterministic ones.

Arguments

Value

a MatrixData.

quantile(alphas,na.rm=FALSE)

Get the cutoffs for each percentile in alphas.

Arguments

alphas	-	A numeric vector with elements between 0 and 1 of percentiles to find cutoffs for.
na.rm	-	A boolean regarding whether to remove NA values before computing the quantiles.

Value

an ArrayData.

undebug(string)

A function for ceasing to debug methods. Normally a method will call the browser command every time it is run. This command will stop it from doing so.

Arguments

string	-	The name(s) of the methods to stop debugging.

See Also

Inherits from : Forecast

Is inherited by : IncidenceForecast

Examples

IncidenceForecast <- R6Class(
  classname = "IncidenceForecast",
  inherit = SimulatedForecast,
  private = list(
    .data = AbstractSimulatedIncidenceMatrix$new()
  ),
  public = list(
    initialize = function(data=SimulatedIncidenceMatrix$new(),forecastTimes=c()){
      if(data$ncol != length(forecastTimes)){
        stop("The number of columns should be the number of times forecasted.")
      }
      private$.forecastMadeTime = now()
      private$.forecastTimes = forecastTimes
      private$.data = data
    }
  ),
  active = list(
    data = function(value){
      private$defaultActive(".data","private",value)
    }
  )
)

HopkinsIDD/ForecastFramework documentation built on Nov. 10, 2019, 2:15 a.m.