On Apr 22, 2011, at 2:07 AM, Laurent Jégou wrote: > Hello, thank you very much for this solution, i'll try to adapt it to my > problem, i think i'll be able to.
Great! > By the way, i wrote about a "sliding-window" weight matrix, it's an > expression used in satellite images processing, esp. convolution filters, > meaning that the main matrix is not confronted to another one of the same > dimensions, but to a much smaller one. The calculus is executed value by > value (pixels on an image), centering the small matrix on each one, hence the > name of sliding window. Yeah, I'm familiar with the concept in general, but I don't see how you can really apply it to Moran's I calculations; the Moran's I value is, by definition, based on calculating the value for every point compared to every other point, and looking at only points within a smaller window risks getting a value that is not representative of the larger context. But if it makes sense to you in your application, then more power to you. :-> Ben Haller McGill University ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
