Greetings. I've been working with the moving window algorithm in Fragstats 4.1, for a landscape ecology study in the Chilean Metropolitan Region. For this purpose, I'm running 10 moving windows varying from 5x5 to 255x255 pixels in a 4682x4715 disturbance binary raster with 30m pixel resolution to measure contagion and composition.
The problem arises when I'm trying to do the process with window sizes over the 1000m. I find quite disturbing that the process takes less than 5 minutes to complete (contrary to window sizes under 1000m, where it takes up to 20 minutes to complete), but the progess bar freezes at the "100%" at the end of the bar.
I tried to split the raster in small sub-landscapes, and to use only one parameter at the time, in order to reduce the memory usage (I thought that would be the problem) but the problem persists.
According to the pre-run of the model, I have 4194303k of available memory, the landscape size and window size, uses 86306k and 197k each.
Any thoughts, help, or advices will be highly appreciated.
Thanks.
