Marker infrastructure

[lgatto]

Currently, there are two ways to encode markers: vectors (legacy) and matrices (new encoding) - see ?markers. Some functions have been updated to deal with both (see issue #52). There are a few possibilities for the future:

• update some functions/methods to deal with vector/matrix markers and assume that the legacy infrastructure/functionality will only deal with vector markers, and new functionality will mainly deal with either matrix markers of both. Update accordingly.
• update all functions/methods to deal with vector/matrix markers
• Refactor everything to use matrix markers - note that there exists already converters, namely mrkVecToMat, mrkMatToVec and friends
• Develop a dedicated markers S4 infrastructure to deal with any implementation in a transparent way (from user point of view)

These are the functions (don't think methods would be caught by this) that have formals that match fcol (including *fcol*).

> ls("package:pRoloc")[which(sapply(xx, function(x) any(grepl("fcol", names(x)))))]
 [1] "addLegend"                "addMarkers"              
 [3] "checkFeatureNamesOverlap" "checkFvarOverlap"        
 [5] "getMarkerClasses"         "getMarkers"              
 [7] "getPredictions"           "isMrkMat"                
 [9] "isMrkVec"                 "knnClassification"       
[11] "knnOptimisation"          "knnOptimization"         
[13] "knntlClassification"      "knntlOptimisation"       
[15] "ksvmClassification"       "ksvmOptimisation"        
[17] "ksvmOptimization"         "markerMSnSet"            
[19] "minClassScore"            "minMarkers"              
[21] "move2Ds"                  "mrkMatAndVec"            
[23] "mrkMatToVec"              "mrkVecToMat"             
[25] "nbClassification"         "nbOptimisation"          
[27] "nbOptimization"           "nnetClassification"      
[29] "nnetOptimisation"         "nnetOptimization"        
[31] "perTurboClassification"   "perTurboOptimisation"    
[33] "perTurboOptimization"     "phenoDisco"              
[35] "plot2D"                   "plot2Ds"                 
[37] "plsdaClassification"      "plsdaOptimisation"       
[39] "plsdaOptimization"        "rfClassification"        
[41] "rfOptimisation"           "rfOptimization"          
[43] "sampleMSnSet"             "showMrkMat"              
[45] "svmClassification"        "svmOptimisation"         
[47] "svmOptimization"          "testMarkers"             
[49] "testMSnSet"               "unknownMSnSet"