RealizationCheck

RealizationCheck(
    names: list,
    statistics: Table,
    probabilities: ndarray,
    quantiles: ndarray | None,
    score_quantiles: ndarray | None,
    data_quantiles: ndarray | None,
    proportions: ndarray | None,
    data_proportions: ndarray | None,
    directions: list | None,
    variograms: list | None,
    data_variograms: list | None,
    models: list | None,
    correlations: ndarray | None,
    data_correlation: ndarray | None,
)

What check_realizations measured, per variable v (category for categorical realizations), realization r and direction d.

Attributes:

Name	Type	Description
names	list of str	Variable names, or category codes as text.
statistics	Table	One row per variable and realization, realization 0 being the declustered data: variable, realization and the statistics of describe, or proportion for categories.
probabilities	(ndarray, shape(p))	Cumulative probabilities at which quantiles are taken.
quantiles, score_quantiles	ndarray, shape (v, r, p), or None	Quantiles of each realization, in data units and in the normal scores of the declustered data, whose target is normal_ppf(probabilities). None for categories.
data_quantiles	ndarray, shape (v, p), or None	Declustered quantiles of the data.
proportions	ndarray, shape (r, v), or None	Share of the nodes in each category, per realization; None for continuous variables.
data_proportions	ndarray, shape (v,), or None	Declustered proportions of the data.
directions	list of (float, float), or None	(azimuth, dip) of each variogram direction; None when omnidirectional.
variograms	list, or None	variograms[v][r][d], the ExperimentalVariogram of realization r; None without lag.
data_variograms	list, or None	data_variograms[v][d], the same for the data.
models	list of Variogram or None	Model each variable was simulated with, when given.
correlations	ndarray, shape (r, v, v), or None	Correlation matrix of each realization, with several variables.
data_correlation	ndarray, shape (v, v), or None	Declustered correlation matrix of the data.