2008-04-03: 1.05-02

	. Bugfixes: misspelled variables and misplaced brackets

2008-04-01:

	. Minor improvements in precision of color alignment in plotHclustColors 

2008-03-26: 1.05-1

	. bugfix: mergeCloseModules behaves properly when input 'colors' is a factor, 
	in particular it drops unused levels to avoid returning empty modules.

2008-03-12: 1.05

	. error trapping in moduleEigengenes revised: hubs can be substituted
	for eigengenes whose calculation failed.

	. mergeCloseModules gets its own trapErrors, but it isn't all that
	useful right now.


2008-03-07: 1.04-01

	. bugfixes in multiSetMEs

2008-03-06: 1.04

	. moduleEigengenes and multiSetMEs now optionally trap errors in module
	eigengene calculation and remove the offending modules from the
	result.

	. moduleEigengenes and multiSetMEs can optionally exclude the grey
	module eigengene from the returned PCs.

	. moduleEigengenes and multiSetMEs can optionally return only the
	valid PCs

(2008-03-03)

	colors2labels checks for NAs in labels

2008-02-27

	1.03-04: colors2labels: return value now has the dimensions of input
	labels, so works correctly on matrices and such.

2008-02-24

	1.03-03: changing default MEprefix to "ME". setMEprefix doesn't seem to work!

2008-02-20

	mergeCloseModules: all print statements are printFlush for consistency.

2008-02-16:
	plotHclustColors now uses names(colors) as rowNames if the latter are NULL and the former
	are valid (have the right lenght).

2008-02-14:
	Fixed a couple of bugs in .clustOrder. 

2008-02-12:
	Fixed problems in mergeCloseModules, consensusOrderMEs, orderMEs when
	number of proper modules is 1 or none.  

2008-02-05:
	Fixed error in removeGreyME.

2008-01-31:
	Added the function removeGreyME

2008-01-29: Version 1.03: 
	Minor robustness improvements: 
	 . mergeCloseModules won't fail if there are less than two proper modules.
	Addition of options:
	 . moduleColor.[set,get]MEprefix: the prefix used to label module
	 eigengenes. For compatibility with old code, standard startup value
	 is "PC", but recommended value is "ME".

2008-01-24: Minor bug fixes:
	Make the call to impute.knn more crash-resistant: The parameter k now doesn't exceed 
	number of supplied genes -1.
	White was listed twice in BaseColors; second instance removed (thanks
	to Jeremy Miller for noticing this)

2007-11-21: Functionality improvement/bug fix:

	checkSets: now accepts parameter useSets to only check selected
	components of the data vector. This fixes a problem with calling
	consensusOrderMEs, mergeCloseModules etc. with multi-data but only for
	one set.

2007-09-18: version: 1.02

	mergeCloseModules: changed default useAbs to FALSE (!!!!)
	                   fixed a bug in counting of old and new modules
			   fixed a bug appearing when no merging was necessary
			   
	plotHclustColors, moduleEigengenes: color changed to colors (!)

	moduleEigengenes: Decreased verbosity for verbose=1

	orderMEs: grey ME not found triggers a warning instead of an error.

	labels2colors: if number of labels exceeds number of availbale colors,
	recycle colors with suffix .1, .2 etc added. Caution: this will make
	extra colors incompatible with standard plotting, but compatible with
	mergeCloseModules (which relies on names of different modules being
	unique).

	normalizeLabels: fixed a bug that merged modules of same size because
			rank was resolving ties by averages.


2007-09-17: version 1.01.01

Bug fixes and small changes:

	moduleEigengenes: fixed problem with saving .Random.seed if the variable doesn't exist (yet)

	multiSetMEs: added argument impute

	plotHclustColors: Fixed to work if color is a single vector of colors

	updated URL in DESCRIPTION
