plot.credpart {evclust} | R Documentation |
Plotting a credal partition
Description
Generates plots of a credal partition.
Usage
## S3 method for class 'credpart'
plot(
x,
X = NULL,
...,
mfrow = c(1, 1),
ytrue = NULL,
Outliers = TRUE,
Approx = 1,
cex = 1,
cexvar = "pl",
cex_outliers = 1.3,
cex_protos = 1,
lwd = 2,
ask = FALSE,
plot_Shepard = FALSE,
plot_approx = TRUE,
plot_protos = TRUE,
xlab = expression(x[1]),
ylab = expression(x[2])
)
Arguments
x |
An object of class |
X |
A data matrix. If it has more than two columns (attributes), only the first two columns are used. |
... |
Other arguments to be passed to the plot function. |
mfrow |
A 2-vector defining the number of rows and columns of the plot. If mfrow=c(1,1), only one figure is drawn. Otherwise, mfrow[1] x mfrow[2] should not be less than x, the number of clusters. |
ytrue |
The vector of true class labels. If supplied, a different color is used for each true cluster. Otherwise, the maximum-plausibility clusters are used instead. |
Outliers |
If TRUE, the outliers are plotted, and they are not included in the lower and upper approximations of the clusters. |
Approx |
If Approx==1 (default), the lower and upper cluster approximations are computed using the interval dominance rule. Otherwise, the maximum mass rule is used. |
cex |
Maximum size of data points. |
cexvar |
Parameter determining if the size of the data points is proportional to the plausibilities ('pl', the default), the plausibilities of the normalized credal partition ('pl.n'), the degrees of belief ('bel'), the degrees of belief of the normalized credal partition ('bel.n'), or if it is constant ('cst', default). |
cex_outliers |
Size of data points for outliers. |
cex_protos |
Size of data points for prototypes (if applicable). |
lwd |
Line width for drawing the lower and upper approximations. |
ask |
Logical; if TRUE, the user is asked before each plot. |
plot_Shepard |
Logical; if TRUE and if the credal partition was generated by kevclus the Shepard diagram is plotted. |
plot_approx |
Logical; if TRUE (default) the convex hulls of the lower and upper approximations are plotted. |
plot_protos |
Logical; if TRUE (default) the prototypes are plotted (for methods generating prototypes, like ECM). |
xlab |
Label of horizontal axis. |
ylab |
Label of vertical axis. |
Details
This function plots different views of a credal partition in a two-dimensional attribute space. If mfrow=c(1,1) (the default), the function plot the dataset with a different symbol for each cluster.
Value
The maximum plausibility hard partition, as well as the lower and upper approximations
of each cluster are drawn in the two-dimensional space specified by matrix X
. If
prototypes are defined (for methods "ecm"
and "cecm"
), they are also
represented on the plot. For methods "kevclus"
, "kcevclus"
or "nnevclus"
a second plot with Shepard's diagram (degrees of conflict vs. transformed dissimilarities) is drawn.
If input X
is not supplied and the Shepard diagram exists, then only the Shepard diagram is drawn.
References
T. Denoeux and O. Kanjanatarakul. Beyond Fuzzy, Possibilistic and Rough: An Investigation of Belief Functions in Clustering. 8th International conference on soft methods in probability and statistics, Rome, 12-14 September, 2016.
M.-H. Masson and T. Denoeux. ECM: An evidential version of the fuzzy c-means algorithm. Pattern Recognition, Vol. 41, Issue 4, pages 1384–1397, 2008.
T. Denoeux, S. Sriboonchitta and O. Kanjanatarakul. Evidential clustering of large dissimilarity data. Knowledge-Based Systems, vol. 106, pages 179-195, 2016.
See Also
extractMass
, summary.credpart
, ecm
,
recm
, cecm
, kevclus
.
Examples
## Example with Four-class data
data("fourclass")
x<-fourclass[,1:2]
y<-fourclass[,3]
c=4
## Running k-EVCLUS with singletons
clus<-kevclus(x=x,k=100,c=c,type='simple')
## Plot the results
plot(clus,X=x,mfrow=c(2,2),ytrue=y)