veri_stats: Verification Statistics
In JuGross/ensAR: Autoregressive postprocessing methods for ensemble forecasts

Description Usage Arguments Details Value Author(s) Examples

Computes different verification statistics for one predictive normal distribution, or for the spread-adjusted linear pool (SLP) of two predictive normal distributions.

1 2	veri_stats(x, par_one = list(mean = NULL, sd = NULL), par_two = par_one, weight_one = NULL, scale = NULL)

`x`	A vector of observations of a weather quantity.
`par_one`	A list with two vectors `mean` and `sd` of the same length as `x` respectively, providing parameters of the normal predictive distribution.
`par_two`	A list with two vectors `mean` and `sd` of the same length as `x` respectively, providing parameters of the second normal predictive distribution. Only required if `weight_one` is not `NULL` in which case the two parameter vectors must be of the same length as `x`.
`weight_one`	A vector of the same length as `x` providing the weights (corresponding to the first normal) used for the spread-adjusted linear pool.
`scale`	A vector of the same length as `x` providing the common scale parameter used for the spread-adjusted linear pool.

If weight_one is NULL verification statistics are computed for a normal distribution with parameters provided by par_one. Otherwise the predictive distribution is the spread-adjusted linear pool (SLP) of two normals with parameters provided by par_one and par_two and SLP parameters weight_one (weights corresponding to par_one) and common scale parameter scale.

A list with elements

pit: A vector of the same length as x giving probability integral transform (PIT) values.
crps: A vector of the same length as x giving the continuous ranked probability score (CRPS) values.
dss: A vector of the same length as x giving the Dawid and Sebastiani score (DSS) values.
pitvar: The sample variance of the PIT values.
rmv: The square root of the mean of the predictive variance.