Home

The package rpgraph allows constructing, manipulating and analysing principal graphs in R. The code is relies on the Java functions written by Andrei Zynoviev available in the java library VDAOEngine. The R code interfaces with the java library via the rJava package. A MATLAB implementation of elastic principal graphs developed and mainteined by Andrei Zynoviev is available here.

An essential overview of Pircipal elastic circles

For an overview of theory behind principal graphs see description provided by Andrei. Further details will be presented here in a future update.

rpgraph prerequisite and installation

The package is currently under heavy development, non very well documented, and only available on GitHub. A functional Java Virtual machine for your system is necessary. Before installing the package, it is advisable to install rJava from sources by using

install.packages(pkgs = "rJava", repos="http://rforge.net", type = 'source')

Compiling from source requires the appropriate development tools, e.g., C/C++ compiler. The installation of the package requires the devtools package, which is available from CRAN. The rpgraph package can be installed using

install.packages("devtools")
library(devtools)
install_github("Albluca/rpgraph")

To take advantage of all the feature of the package, it is advisable to also install (not necessary from source in this case) bigpca, flashpcaR, irlba, nsprcomp, and plotly.

flashpcaR is available from GitHub

library(devtools)
install_github("gabraham/flashpca")

The other packages can be installed from CRAN.

install.packages(c("bigpca", "flashpcaR", "irlba", "nsprcomp", "plotly"))

Workaround for common problems

The installation and loading of rJava is known to be problematics under certain circumstances on MacOS. A number of workaround can be found on the internet. Possible solutions include recompiling the package, setting certain environment variables manually, ~ and changing the operating system~. Most of the times the problem is connected with the package struggling to find the appropriate information in the global environment.

If rJava fails to load try typing

options("java.home"="/Library/Java/JavaVirtualMachines/jdk1.7.0_79.jdk/Contents/Home/jre")
dyn.load('/Library/Java/JavaVirtualMachines/jdk1.7.0_79.jdk/Contents/Home/jre/lib/server/libjvm.dylib')

before loading the library. These lines may need to get adjusted depending on the version of the virtual machine that the user intendt to use.

Basic functionalities

The basic functionality of rpgraph is provided by the computeElasticPrincipalGraph function, which can be used to compute a list of principal graphs. For each principal graph is then possible to obtain debug information (plotMSDEnergyPlot and accuracyComplexityPlot) and to plot the result in 2 and 3 dimensions (plotData2D, plotPieNet, plotData3D). The principal graph can also be converted into a igraph network (ConstructGraph).

By defaut, the principal graph is constructed from an empty set of nodes. However, it is also possible to specify an initial configuation by specifying a set of nodes and edges among them. This allows, not only to observe how the principal graoh is created step-by-step, but also to combine the different grammars to obtain more coplex topologies.

The points a set of points can be projected on the principal graph with the same number of dimensions in two different ways. Using TaxonList it is possible to obtains a list of points associated with each node of the graph, while using projectPoints is it possible to project the points on the edges. Note that, expecially in high dimensional spaces, it is possible tha a point is projected on an nodes instead of an edge. More details on this will be available in future updates.

Once the points have been projected on the edges, it is possible to order them on a path using the fucntion OrderOnPath. This allows for the derivation of a pseudo time across the points.