conditionalReordering: Conditional Reordering

Description Usage Arguments Value

View source: R/reordering.R


function to generate ranks that have been simply reordered with a Gaussian copula or conditionally reordered with Gaussian copula stressed scenarios from a base Gaussian copula.


conditionalReordering(n, list.correlation.matrix, name,
  scenario.probability = NULL, region.boundaries = NULL,
  region.probability = NULL, keep.realized.scenario = F)



positive numeric value of length one. The number of ranks to produce (equal to the number of simulations of the model).


list of correlation matrices, the correlation matrix corresponding to the base normal copula should be provided as a named member "base" in the list (and in first position). the rest of the scenarios should be named in the list by a unique identifier that should match the column names of the argument region.boundaries. Please consider that if no scenario correlation matrices are provided, then simple reordering with the "base" correlation matrix is undertaken (note also that in this case, we require scenario.probability, region.boundaries and region.probability to be NULL).


character value of length between 0 and 4. It should indicate the names of the subset of risks among:

  • market

  • life

  • health

  • nonlife

that are aggregated together with the reordering algorithm. The order of risks in this vector should respect the order defined in the correlation matrices in list.correlation.matrix.


numeric value giving the scenario probabilities (these probabilities should be provided in the same order as the the order of scenarios in list.correlation.matrix (following the correlation matrix named "base").


matrix with named columns and rows giving the thresholds for each regions (boundaries of the scenario rectangles). Each line represents a given scenario and each column a given quantity to reorder. The rownames should match the scenario names and the colnames should match the risks respecting the order prescribed in both name and the colnames of each correlation matrix in list.correlation.matrix.


numeric vector giving the probability under the base Gaussian copula (characterized by the correlation matrix named "base") to hit the scenario regions given by each line in regions.boundary.


logical value. Should we keep the realized scenario for each line?


a data.table with the final ranks (between 0 and 1) with which we should reorder the given simulations.

sstModel documentation built on May 4, 2018, 1:04 a.m.