[Robast-commits] r998 - branches/robast-1.1/pkg/RobExtremes/inst/AddMaterial

[noreply at r-forge.r-project.org]

Author: ruckdeschel
Date: 2018-07-20 14:33:17 +0200 (Fri, 20 Jul 2018)
New Revision: 998

Modified:
   branches/robast-1.1/pkg/RobExtremes/inst/AddMaterial/getLMPareto.R
Log:
[RobExtremes] branch 1.1 added explanation of scale in-/equivariance for Pareto to code for the computation of LMs for Pareto

Modified: branches/robast-1.1/pkg/RobExtremes/inst/AddMaterial/getLMPareto.R
===================================================================
--- branches/robast-1.1/pkg/RobExtremes/inst/AddMaterial/getLMPareto.R	2018-07-20 12:27:03 UTC (rev 997)
+++ branches/robast-1.1/pkg/RobExtremes/inst/AddMaterial/getLMPareto.R	2018-07-20 12:33:17 UTC (rev 998)
@@ -1,4 +1,14 @@
 ### getLMs for Pareto
+### produced Jul 20, 2018 P.Ruckdeschel
+
+## note:: the only param to be estimated in the smooth model framework
+##        is the shape k
+##        now, independent of fixed parameter Min
+#         L2deriv in this case is distributed as (1-Exp(1))/k
+##        so shape k acts as a scaling parameter, and we can compute
+##        LMs under reference value k = 1, Min = 1,
+##        the respective for general k OptICs then are simply  k \times OptIC(k=1)
+
 .checkIC <- function(IC,L2deriv,Distr){
    IC. <- as(diag(dimension(IC at Curve)) %*% IC at Curve, "EuclRandVariable")
    IC.. <- function(x) sapply(x,function(y) c(IC. at Map[[1]](q.l(Distr)(y))))