News for Package 'systemicrisk'

Changes in version 0.4.2

• Bug fix for matrices of dimension > 1024.

• Added specific support for some non-square matrices: Model.lambda.constant.nonsquare, Model.p.constant.nonsquare, calibrate_ER.nonsquare.

Changes in version 0.4.1

• Addjusted that non-square matrices can be used (using fixed matrices for p and lambda). See the new vignette as an example.

Changes in version 0.4

• Added methods to calibrate models to a given density.

Changes in version 0.3.1

• Fixed parameter range checks for Fitness models.

Changes in version 0.3 (2015-12-21)

• Added Hierarchical Models.

• Arbitrary values of the Matrix can now be force to equal 0 (not just the diagonal). An error message will be produced if no matrix exists for the given row and column sums. The function findFeasibleMatrix() supersedes getfeasibleMatrix()

• Added automatic choice of thinning to achieve desired relative effective sample size (choosethin)

• Added possibility to fix elements of L to some values.

• Added a fitness based model for the matrix that incorporates power laws.

Changes in version 0.2 (2015-03-20)

• First CRAN upload.