A S4 class to store results from an invariant coordinate system transformation or independent component computation based on two scatter matrices and two location vectors.
Objects can be created by calls of the form new("ics2", ...). But usually objects are created by the function ics2.
The Class inherits from the ics class.
gSkew:Object of class "numeric". Gives the generalized skewness measures of the components
gKurt:Object of class "numeric". Gives the generalized kurtosis measures of the components
UnMix:Object of class "matrix". The unmixing matrix.
S1:Object of class "matrix". The first scatter matrix.
S2:Object of class "matrix". The second scatter matrix.
T1:Object of class "numeric". The first location vector.
T2:Object of class "numeric". The second location vector.
S1name:Object of class "character". Name of the first scatter matrix.
S2name:Object of class "character". Name of the second scatter matrix.
S1args:Object of class "list". Additional arguments needed when calling function S1.
S2args:Object of class "list". Additional arguments needed when calling function S2.
Scores:Object of class "data.frame". The underlying components in the invariant coordinate system.
DataNames:Object of class "character". Names of the original variables.
StandardizeB:Object of class "character". Names standardization method for UnMix.
StandardizegKurt:Object of class "logical". States wether the generalized kurtosis is standardized or not.
For this class the following generic functions are available: print.ics2, summary.ics2
But naturally the other methods like plot, coef, fitted and so from class ics work via inheritance.
Klaus Nordhausen
ics2