Changes in version 1.10:

USER VISIBLE CHANGES

    o   Using NEWS.Rd

    o   batch slot in CNSet objects is class 'character' (previously, class
	 was 'factor')

    o   the ff package is required for preprocessing and genotyping prior to
	 copy number analyses with the crlmmCopynumber function.

    o   We have added several vignettes pertaining to copy number analyses
	 in the crlmm package: CopyNumberOverview, AffymetrixPreprocessCN,
	 IlluminaPreprocessCN, and Infrastructure.  The
	 AffymetrixPreprocessCN and IlluminaPreprocessCN vignettes provide
	 instructions for preprocessing and genotyping the raw intensities
	 for Affymetrix and Illumina platforms, respectively.  The
	 'copynumber' and 'Infrastructure' vignettes are applicable to both
	 the Illumina and Affymetrix platforms. The CopyNumberOverview
	 vignette provides a brief summary of the available vignettes for
	 copy number analysis.

    o   For those interested in copy number analysis of Illumina platforms,
	 we added functions that break-down the preprocessing and genotyping
	 steps: constructInf, preprocessInf, and genotypeInf. The
	 genotype.Illumina function is now a wrapper for these functions.

    o   additional documentation for crlmm is provided in a compendium:
	 http://www.biostat.jhsph.edu/~rscharpf/crlmmCompendium/index.html

Changes in version 1.8:

USER VISIBLE CHANGES

    o   define CNSet class with batch and batchStatistics slots

    o   deprecate CNSetList class

    o   deprecate crlmmWrapper, genotype2, and crlmmCopynumber2.  Use
	 crlmmCopynumber in place of crlmmWrapper and crlmmCopynumber2.  Use
	 genotype instead of genotype2.

    o   raw copy number estimates no longer stored in container.  CA and CB
	 methods are no longer simple accessors to data stored in the
	 eset-extension.  These functions compute raw copy number estimates
	 from the linear model parameters.

Changes in version 1.3:

USER VISIBLE CHANGES

    o   3 new classes created:
	
	  * 'ABset': container for quantile-normalized A and B intensities
	     for both SNP and copy number probes.  Required assay data
	     elements are 'A' and 'B'.  Extends eSet directly.
	
	      1. For nonpolymorphic probes, the quantile normalized
	         intensity is stored in the 'A' assay data element.  The
	         corresponding row in the 'B' assay data element is NA.
	         This is a bit inefficient, but greatly simplifies
	         downstream analyses.  In particular, '[' works.
	
	  * 'CrlmmSetList': container for results from preprocessing and
	     genotyping.  This object is a list.  The first element of the
	     list is an ABset.  The second element is a SnpSet containing
	     genotype calls.  The two elements are required to have
	     identical featureNames and sampleNames.
	
	      1. added several methods for subsetting and accessing elements
	         of this object, including featureNames, sampleNames, and
	         '['.
	
	  * 'CopyNumberSet': contains locus-level estimates of copy number
	     for SNPs and polymorphic probes.
	
	      1. Required assay data elements are 'CA' and 'CB',
	         corresponding to the absolute copy number for allele A and
	         B, respectively.
	
	      2. For nonpolymorphic probes, the total copy number is stored
	         in the 'CA' slot and a NA is recorded for the corresponding
	         row in the CB matrix.
	
	      3. Useful methods: 'copyNumber', 'ellipse', 'points'

    o   'crlmmWrapper' function does preprocessing (quantile-normalization)
	 and genotyping, saving an object of class CrlmmSetList for each
	 chromosome

    o   'computeCopynumber' now requires an object of class 'CrlmmSetList'
	 and returns an object of class 'CopyNumberSet'.