[R-sig-Geo] Space-time cluster analysis

[Virgilio Gómez-Rubio]

Hi Joe,


> I am using space-time cluster analysis to search for areas/times of high
> disease incidence.  My data is aggregated by geographic areas with known
> background population numbers.  So under the null hypothesis, the number of
> cases in a given location/time period is Poisson distributed, dependent on
> the number of person-years.

Some of the code in DCluster and spatclus may be adapted to detect
clusters in space and time. I am planning to add a space-time version of
the spatial scan statistic but there is no code at the moment. However,
it should not be difficult to implement. Basically, you need to
aggregate data over a temporal window and then perform a spatial test.

Hope this helps.

Virgilio