AnthropMMD (version 4.0.3)

plot.anthropmmd_result: Display a multidimensional scaling (MDS) plot with the MMD dissimilarities as input

Description

This function plots a 2D or 3D MDS to represent the MMD dissimilarities among the groups compared. Various MDS methods are proposed, and most of them are based on the R package smacof.

Usage

# S3 method for anthropmmd_result
plot(x, method = c("classical", "interval", "ratio", "ordinal"),
axes = FALSE, gof = FALSE, dim = 2, asp = TRUE, xlim = NULL, ...)

Value

This function returns no value by itself, and only plots a MDS in a new device.

Arguments

x

An object of class anthropmmd_result, produced by the function mmd.

.

method

Specification of MDS type. classical uses the metric MDS implemented in stats::cmdscale; the three other values are passed to the R function smacof::smacofSym (see its help page for more details).

axes

Boolean: should the axes be displayed on the plot?

gof

Boolean: should goodness of fit statistics be displayed on the topleft corner of the plot? More details below.

dim

Numeric value, 2 or 3. Indicates the maximal dimension desired for the MDS plot. It should be noted that, even with dim = 3, the final solution may include only two axes.

asp

Boolean. If TRUE, the same scale is used for all axes. More details below.

xlim

Parameter passed to plot, can be NULL.

...

Other arguments possibly passed to plot.

Author

Frédéric Santos,

Details

• Axes and scale. Making all axes use the same scale is strongly recommended in all cases (Borg et al., 2013). For a 3D-plot, since the third axis carries generally only a very small percentage of the total variability, you might want to uncheck this option to better visualize the distances along the third axis. In this case, the axes scales must be displayed on the plot, otherwise the plot would be misleading.

• Goodness of fit values. (i) For classical metric MDS, a common statistic is given: the sum of the eigenvalues of the first two axes, divided by the sum of all eigenvalues. It indicates the fraction of the total variance of the data represented in the MDS plot. This statistic comes from the $GOF value returned by the function stats::cmdscale. (ii) For SMACOF methods, the statistic given is the $stress value returned by the function smacof::smacofSym It indicates the final stress-1 value. A value very close to 0 corresponds to a perfect fit. (iii) For both approaches, a 'rho' value is also given, which is the Spearman's correlation coefficient between real dissimilarities (i.e., MMD values) and distances observed on the MDS plot (Dzemyda et al.,2013). A value very close to 1 indicates a perfect fit.

References

G. Dzemyda, O. Kurasova and J. Zilinskas (2013) Multidimensional Data Visualization, Springer, chap. 2, p. 39--40.

I. Borg, P. Groenen and P. Mair (2013) Applied Multidimensional Scaling, Springer, chap. 7, p. 79.

See Also

start_mmd, stats::cmdscale, smacof::smacofSym

Examples

## Load and visualize a binary dataset:
data(toyMMD)
head(toyMMD)

## Convert this dataframe into a table of sample sizes and relative
## frequencies:
tab <- binary_to_table(toyMMD, relative = TRUE)
tab

## Compute and display a symmetrical matrix of MMD values:
mmd_out <- mmd(tab, angular = "Freeman")

## Plot a classical metric MDS in two dimensions:
plot(x = mmd_out, method = "classical",
     axes = TRUE, gof = TRUE, dim = 2)