ens_brier: Brier score and its decomposition for an ensemble.

In andrew-MET/harpPoint: Point verifition for NWP forecasts

Brier score and its decomposition for an ensemble.

Description

Brier score and its decomposition for an ensemble.

Usage

ens_brier(
  .fcst,
  parameter,
  thresholds,
  groupings = "leadtime",
  climatology = "sample",
  rel_probs = NA,
  num_ref_members = NA,
  keep_score = c("both", "brier", "reliability"),
  show_progress = FALSE
)

Arguments

.fcst	A harp_fcst object with tables that have a column for observations, or a single forecast table.
parameter	The name of the column for the observed data.
thresholds	A numeric vector of thresholds for which to compute the Brier Score.
groupings	The groups for which to compute the ensemble mean and spread. See group_by for more information of how grouping works.
climatology	The climatology to use for the Brier Skill Score. Can be "sample" for the sample climatology (the default), a named list with elements eps_model and member to use a member of an eps model in the harp_fcst object for the climatology, or a data frame with columns for threshold and climatology and also optionally leadtime.
rel_probs	Probabilities to use for reliability diagrams. Set to NA (the default) to select automatically.
num_ref_members	The number of members for which to compute the fair Brier score.
keep_score	ens_brier computes the Brier Score and its components, as well as reliability. If you only want to output the Brier score and its decomposition set to "brier", to only keep the reliability information set to "reliability". To keep both, set to "both". The default behaviour is "both".
...	Used internally depending on the class of the input.

Value

A data frame with data grouped for the groupings column(s) and columns for brier_score, brier_skill_score and the deomposition of the brier score - brier_score_reliability, brier_score_resolution and brier_score_uncertainty.

andrew-MET/harpPoint documentation built on Feb. 23, 2023, 1:06 a.m.