The curse of dimensionality is a blessing in disguise. Although multi-dimensional arrays in R are restricted to the six basic types, many programming languages vectors and arrays as a template, which means that you can make such compound objects from any type of object that you like.
This becomes especially handy when it comes to organisation of high-dimensional, spatial data.
The base class
VirtualArray is defined, which acts as a wrapper around lists allowing users to fold arbitrary sequential data into n-dimensional, R-style virtual arrays. The derived
XArray class is defined to be used for homogeneous lists that contain a single class of objects. The
SfArray classes enable the use of stacked spatial data instead of lists.
The functions here were originally developed and published as part of the chronosphere R package. For better compliance with UNIX philosophy and more efficient distribution/development, the original chronosphere has been broken up to three R packages:
rgplates: functions related to tectonic reconstructions.
via: Virtual Arrays for efficient organisation of high-dimenional data.
chronosphere: version-controlled data distribution.
This is a pre-alpha version, and like R, comes with absolutely no warranty.