EnsRps: Calculate the ensemble-adjusted Ranked Probability Score (RPS) for categorical forecasts
Description
Calculate the ensemble-adjusted Ranked Probability Score (RPS) for categorical forecasts
Usage
EnsRps(ens, obs, R.new = NA, format = c("category", "members"))
FairRps(ens, obs, format = c("category", "members"))
Arguments
ens
matrix with N rows representing N time instances of categorical ensemble forecasts as follows: If `format = category` (the default), then ens[t,r] indicates the category that the r-th ensemble member predicts for time t. Note that categories must be positive integers. If `format = members`, then ens[t,k] is the number of ensemble members that predict category k at time t.
obs
vector of length N, or matrix with N rows, representing the N observed category as follows: If `format = category', obs is a vector and obs[t] is the category observed at time t. If `format = members`, obs is a matrix where obs[t,k] = 1 (and zero otherwise) if category k was observed at time t
R.new
ensemble size for which the scores should be adjusted, defaults to NA (no adjustment)
format
string, 'category' (default) or 'members' (can be abbreviated). See descriptions of arguments `ens` and `obs` for details.
Value
numeric vector of length N with the ensemble-adjusted RPS values