fwdmv.object {Rfwdmv} | R Documentation |
fwdmv.object
Description
An object containing a fitted forward search on multivariate data. The class attribute is set to fwdmv
.
Format
An fwdmv.object
is a list with the following elements:
- call
- the matched call.
- Distances
- a numeric matrix containing the Mahalanobis distances computed during the forward search.
- Center
- a list of numeric matrices containing the location estimates for each group computed during the forward search.
- Cov
- a list of numeric matrices containing the covariance matrix estimates (in packed storage) for each group computed during the forward seach).
- Determinant
- a list of numeric vectors containing the determinants of the covariance matrix estimates for each group computed during the forward search.
- Unit
- a matrix with one row for each unit in the data set and one column for each step in the forward search. The (i,j) element is 1 if unit i is in the subset during step j of the forward search and is 0 otherwise.
- groups
- a list of integer vectors containing the user specified tentative groups.
- n
- an integer, the number of units in the data.
- p
- an integer, the number of variables in the data.
- m
- an integer, the number of units in the subset during the first step of the forward search.
- data
- a numeric matrix containing the data, the dimnames attribute is set to
NULL
.
- data.name
- the name of the data frame or matrix containing the multivariate data set.
- data.names
- a list of character vectors containing the row and column names of the data.
- group.names
- a character vector containing the names of the tentative groups.
- unassigned
- an integer vector containing the indices of the units that do not belong to one of the tentative groups.
- constrained
- a loigcal value,
TRUE
if the forward search was constrained.
- scaled
- a logical value,
TRUE
is scaled Mahalanobis distances were used during the forward search.
- Max
- a numeric vector contining the maximum Mahalanobis distance in the subset.
- Mth
- a numeric vector containing the mth overall Mahalanobis distance.
- Min
- a numeric vector containing the minimum Mahalanobis distance not in the subset.
- Mpo
- a numeric vector containing the (m+1)th overall Mahalanobis distance.
- initial
- a logical value,
TRUE
if this object was generated by the function fwdmv.init
.
Details
Class fwdmv
objects are created by the functions fwdmv
, fwdmv.init
, and partition
. The Rfwdmv
package contains a variety of plot methods for assesing fwdmv
objects. These methods are listed in the see also section. Note that there are accessor methods for several elements contained in the fwdmv
object. When an accessor method exists for a certain element it should be used to retrieve that element in preference to direct reference as the structure of the fwdmv
object is likely to change.
See Also
fwdmv.init
- fit an ititial multivariate forward search.
fwdmv
- fit a multivariate forward search with user specified tentative groups.
Plot methods for assessing a fitted forward search stored in an fwdmv
object:
fwdmvPairsPlot
- a pairs-like plot.
fwdmvQuantilePlot
- plot trajectories of entering units over quantiles of the distances in the subset.
fwdmvEllipsePlot
- a pairs-like plots with the subsets represented by ellipses.
fwdmvConfirmPlot
- plots the nearest center for unassigned and misclassified units.
fwdmvCovariancePlot
- a forward plot of the elements of the covariance matrices.
fwdmvDeterminantPlot
- a forward plot of the determinants of the covariance matrices.
fwdmvDistancePlot
- a forward plot of the Mahalanobis distances.
fwdmvEccentricityPlot
- a forward plot of the eccentricity for one biariate ellipse.
fwdmvPrincompPlot
- a forward plot of the principal components.
fwdmvEigenvectorPlot
- a forward plot of a user specified eigenvector of the covariance matrice.
fwdmvEntryPlot
- a forward entry plot.
fwdmvGapPlot
- a gap plot.
fwdmvMinmaxPlot
- minimum and maximum distances plot.
fwdmvChangePlot
- aa forward plot of change in Mahalanobis distance.
partition
- graphically assign units to groups.
fwdmvPartitionPlot
- view tentative group.
Accessor methods:
eigenvalues.fwdmv
- the eigenvalues of the covariance matrix estimates computed during the forward search.
eigenvectors.fwdmv
- the eigenvectors of the covariance matrix estimates computed during the forward search.
Examples
data(fondi.dat)
fondi.1 <- fwdmv(fondi.dat)
#fondi.1 is an fwdmv object.
[Package
Rfwdmv version 0.72-2
Index]