I'm an intermediate-ish user of ArcGIS 10.2 (win7 pro, hardware robust enough to handle very large datasets and computationally intensive analysis) with perhaps a simple problem that's evading me right now.
I have a raster dataset with different classes of land values. I am interested in whether a dichotomous variable corresponding to presence/non-presence of a land classification is spatially clustered within different zones of a polygon feature. Basically I'd like to do something similar to tabulate area or zonal statistics, whereby I could obtain Moran's I for each individual zone in the feature class. To start I'd convert the raster to a polygon, but based on what I'm seeing I would have to compute Moran's I over the entire overlapping polygon with the 'zones,' rather than obtaining values for each individual one. (I could do it one by one, but this seems cumbersome.)