statistics.predictive: Summary statistics from predictive distributions
In geoR: Analysis of Geostatistical Data

statistics.predictive

R Documentation

Summary statistics from predictive distributions

Computes summaries based on simulations of predictive distribution returned by krige.bayes and krige.conv.

statistics.predictive(simuls, mean.var = TRUE, quantile, threshold,
                      sim.means, sim.vars)

`simuls`	object with simulations from the predictive distribution
`mean.var`	Logical. Indicates whether or not to compute mean and variances of the simulations at each location.
`quantile`	defines quantile estimator. See documentation for `output.control` .
`threshold`	defines probability estimator. See documentation for `output.control`.
`sim.means`	Logical. Indicates whether or not to compute the mean of of the conditional simulations.
`sim.vars`	Logical. Indicates whether or not to compute the variances of the conditional simulations.

A list with one ore more of the following components.

`mean`	mean at each prediction location.
`variance`	variance at each prediction location.
`quantiles`	quantiles, at each prediction location.
`probabilities`	probabilities, at each prediction location, of been below the provided threshold.
`sim.means`	vector with means of each conditional simulation.
`sim.vars`	vector with variances of each conditional simulation.