Pearson correlation in raster datasets

Discussion created by alanfer17 on Sep 27, 2010
Latest reply on Dec 27, 2011 by sarah192837
Dear all,

I need to make a correlation between two raster datasets. I want to use Raster Calculator, but I'm still finding a bit difficult to work it out. I have the formula from statistics books, and I know that I'm required to multiply, and add the grids, right? But it's not so straight forward.
How am I supposed to find N, for instance? Should I just multiply the Columns and Rows from my grid to see how many pixels I have on the dataset?
The formula for the Pearson correlation coefficient is quite long and I don't think I should attempt to do everything in one go. I reckon I should do it by different stages.
Any ideas would be very welcome!

Just a couple more things about the two datasets: one of them is a NDVI that I acquired from LANDSAT 5TM, and the other raster has been originated from an interpolation of points (rainfall gauges). I have used Mask on both grids, therefore they have the same number or rows and columns (same number of pixels?).

Thanks for your time.