The MWA package performs an analysis that I have developed in collaboration with Karsten Donnay. It combines different geospatial and statistical techniques to find out whether geocoded events of a specific type causally affect the numbers of nearby subsequent events of a different type. This comes in very handy if one wants to understand, for example, civilian reactions to insurgent violence in civil conflicts, or the effectiveness of crime prevention measures in urban areas.
You can download the latest version of the software as a package for the
R programming language here here or simply do an
A corresponding paper has been published in Political Geography and the preprint is available here: [pdf].