LOG of the CHANGES in the package ICS

Version 1.4-1

* fixing a mistake in the documentation
* bugfix in summary() method of class ICS so that it doesn't call print()

Version 1.4-0

* changing the maintainer email address and welcoming new contributors to the package
* complete new implementation of ICS that uses S3 classes and methods and that supports different algorithms. From now on function ICS should be used and functions ics and ics2 are to be avoided.

Version 1.3-2

* changing the maintainer email address
* removing latin1 encoding
* updating the citation file

Version 1.3-1

* changing the maintainer email address

Version 1.3-0

* adding new function ics2 and the corresponding print and summary methods
* making the screeplot function a method and improving it
* improving the code of the ics function
* general revision of the help files
* new functions MeanCov and Mean3Cov4

Version 1.2-5

* changing maintainer email address
* cleaning the depends and imports
* fixing some vignette issues
* some minor code improvements

Version 1.2-4

* changing the test statistic in mvnorm.kur.test to make the test affine invariant, thanks to Sara Taskinen
* dropping the internal check if a matrix is numeric since that is quite time consuming
* some minor code improvements
* changing the maintainer email address

Version 1.2-3

* fixing some problems related to R2.15.0
* updating some help files

Version 1.2-2

* added the function scovq
* updating of references

Version 1.2-1

* added the function cov4.wt
* updating of references

Version 1.2-0

* added a vignette to the package
* changed to the citation. Now the correct citation of the package is a reference to the corresponding JSS paper.

Version 1.1-3:

* The covariance matrix for the mahalabobis distance in mean3 is now devided by n instead of n-1.

Version 1.1-2:

* Fixing a bug in cov4 (reported by Sara Taskinen).

Version 1.1-1:

* Minor changes in documentation.

Version 1.1-0:

* Changing the package title description
* Adding more slots to the ICS class and renaming some
* Giving the option to submit also 2 scatter matrices to the function ics.
  This changes the way of the computation method and leads to a new option on how to
  standardize the Kurtosis values. Note that code for ics of version 1.0-0 might break.
* New function tM
* New function covAxis
* Updating some references
* Some new examples and changing examples so, that X %*% A is now X %*% t(A)

Version 1.0-0:

* First version released on CRAN