cmdsr: Continuous embeddings using cmdsr
In ginagruenhage/cmdsr: continuous Multi-dimensional Scaling

The package cmdsr provides functions to compute and analyze continuous embeddings. Continuous embeddings are an extension of Multi-Dimensional Scaling (MDS). They are the solution of multiple MDS problems that are solved simultaneously together with a smoothing penalty. This ensures that the individual MDS solutions are similar local minima, so that they can be connected smoothly. More information about each function can be found in the function documentation.

Computing continuous embeddings

The function cmds computes the continuous embedding. Its only obligatory parameter is the list of distance matrices DL. The most commonly used optional parameters are k, the dimension of the MDS embedding and l, the smoothing parameter.

Plotting continuous embeddings

The function plot.cmds can be used to plot the results from cmds. plot.cmds(res) plots the continuous embedding. It automatically detects the embedding dimension. Setting the optional parameter shepard to TRUE yields a Shepard plot, a common diagnostic tool for MDS. A plot of the convergence trace is computed when calling plot.cmds with convergence = TRUE. Two plots are created. The second is similar to the first, but skips the first iteration. This results in better scaling of the y-axis.

Summarizing continuous embeddings

The function summary.cmds gives a summary of the cmds results.

ginagruenhage/cmdsr documentation built on May 17, 2019, 4:20 a.m.