findModelCluster-class {svcR} | R Documentation |
reading data, computation of the svc model and cluster labelling
Objects can be created by calls of the form new("findModelCluster", ...)
.
or by calling the findModelCluster
function.
WVectorsYA
:"list"
contains
the lagrange coefficients : WVectorsYA$A Matrice
:"list"
contains
variables names Matrice$var, attributes names Matrice$Att and data Matrice$MatMatriceK
:"vector"
contains
kernel matrixData
:"matrix"
contains
Data Matrix MinMaxXY
:"vector"
contains
min max values for first and second coordinatesMisClass
:"vector"
contains
missclassfied pointsDName
:"character"
contains
the prefix name of data for files decodingfileIn
:"character"
contains
data file pathClassPoints
:"vector"
contains
class of grid pointsCx
:"numeric"
contains
the x column id of data matrixCy
:"numeric"
contains
the y column id of data matrixNu
:"numeric"
contains
nu value of the svc modelKNN
:"numeric"
contains
knn value for labellingSizeGrid
:"numeric"
contains
the size grid for labellingAroundNullVA
:"numeric"
contains
almost null value for lagrange coefficient estimationNumPoints
:"vector"
contains
value fo grid pointsNicolas Turenne - INRA France nicolas.turenne@jouy.inra.fr
## exemple with iris data MetOpt = 1; # optimisation method with randomization MetLab = 1; # grid labelling KChoice = 1; # 0: eucli 1: radial 2: radial+dist Nu = 1.0; q = 2000; # lot of clusters K = 1; # only 1 nearest neighbour for clustering Cx = Cy = 0; # we use principal component analysis factors G = 13; # size of the grid for cluster labelling DName = "iris"; fileIn = ""; # fileIn migth be such as "D:/R/library/svc/", if NULL it will work on iris data fmc = findModelCluster( as.integer(MetOpt), MetLab, KChoice, Nu, q, K, G, Cx, Cy, DName, fileIn); fmc@NumPoints; # values of the grid