plot.homals: Plot methods for homals

In homals: Gifi Methods for Optimal Scaling

Description

These methods provide various 2D and 3D plots for objects of class "homals".

Usage

## S3 method for class 'homals'
plot(x, plot.dim = c(1, 2), plot.type = "loadplot", var.subset, main, 
type, xlab, ylab, xlim, ylim, leg.pos = "topright", identify = FALSE, ...)

## S3 method for class 'homals'
plot3d(x, plot.dim = c(1, 2, 3), plot.type = "jointplot", var.subset, type, xlab, ylab, 
zlab, col, main, sphere = TRUE, bgpng = "particle.png", ax.grid = TRUE, ...)

## S3 method for class 'homals'
plot3dstatic(x, plot.dim = c(1, 2, 3), plot.type = "jointplot", var.subset, main, type, 
xlab, ylab, zlab, xlim, ylim, zlim, ...)

Arguments

x	Object of class "homals"
plot.dim	Vector with dimensions to be plotted against.
plot.type	String indicating which type of plot to be produced: "jointplot", "catplot" "labplot", "objplot", "starplot", "graphplot", "hullplot", "lossplot", "prjplot", "spanplot", , "trfplot", "vecplot", "vorplot", "loadplot", "screeplot", "dmplot" (Not all of these are provided as 3d-plots, see below).
var.subset	Numeric vector for subsetting variables to be plotted. If missing, all variables are taken into account. Ignored for non separate variable plots.
main	Plot title.
type	Type of points/lines to be plotted.
xlab	Label of x-axis.
ylab	Label of y-axis.
zlab	Label of z-axis.
col	Plotting color.
xlim	Limits for x-axis.
ylim	Limits for y-axis.
zlim	Limits for z-axis.
leg.pos	Position of the legend (for "trfplot" and "spanplot" only) as provided in legend.
identify	If TRUE, interactive plots are provided.
sphere	Whether 3D-sphere should be drawn.
bgpng	Background image in /library/rgl/textures/. If NULL, blank background.
ax.grid	If TRUE grid is drawn.
...	Further plot arguments passed: see plot, plot3d in package rgl, and scatterplot3d in package scatterplot3d for detailed information.

Details

Plot description:

- Object plot (plot.type = "objplot"): Plots the scores of the objects (rows in data set) on two or three dimensions.

- Category plot (plot.type = "catplot"): Plots the rank-restricted category quantifications for each variable separately. Three-dimensional plot is available.

- Voronoi plot (plot.type = "vorplot"): Produces a category plot with Voronoi regions.

- Joint plot (plot.type = "jointplot"): Plots the object scores and the category quantifications in 1 device. Three-dimensional version is available.

- Graph plot (plot.type = "graphplot"): Joint plot with connections between scores/quantifications.

- Hull plot (plot.type = "hullplot"): For each single variable the object scores are mapped onto two dimensions and the convex hull for each response category is drawn.

- Label plot (plot.type = "labplot"): Similar to object plot, the object scores are plotted but for each variable separately with the corresponding category labels. A three-dimensional version is provided.

- Span plot (plot.type = "spanplot"): As label plot, it maps the object scores for each variable and span plot connects them by the shortest path within each response category.

- Star plot (plot.type = "starplot"): Again, the object scores are mapped. In addition these points are connected with the category centroid. Three-dimensional version is provided.

- Loss plot (plot.type = "lossplot"): Plots the rank-restricted category quantifications against the unrestricted for each variable separately.

- Projection plot (plot.type = "prjplot"): For variables of rank 1 the object scores (two-dimensional) are projected onto a straight line determined by the rank restricted category quantifications.

- Vector plot (plot.type = "vecplot"): For variable of rank 1 the object scores (two-dimensional) are projected onto a straight line determined by the rank restricted category quantifications.

- Transformation plot (plot.type = "trfplot"): Plots the original (categorical) scale against the transformed (metric) scale on each dimension over the categories of each variable separately.

- Loadings plot (plot.type = "loadplot"): Plots the loadings of the variables and connects them with the origin. Three-dimensional version is available.

- Scree plot (plot.type = "screeplot"): Produces a scree plot based on the eigenvalues.

- Discrimination measures (plot.type = "dmplot"): Plots the discrimination measures for each variable.

See Also

homals

Examples

##Graphplot for Hartigan solution: 3 dimensions extracted, dimension 1 plotted
##against dimension 3.
data(hartigan)
res <- homals(hartigan, ndim = 3)
plot(res, plot.dim = c(1,3), plot.type = "graphplot")

homals documentation built on May 2, 2019, 6:51 p.m.