I have a non-normal count variable y and 6 rasters x1 to x6, each mapped onto a (n x n) grid. I have applied a zero-inflated negative binomial regression using the pscl package of R. I now need to generate a risk map using a raster calculator. In this paper, the authors weight their maps after a Poisson regression using the expression
Exp(\beta_0+\sum(\beta_i * x_i))/Exp(\beta_0)
In the current case I have two additional parameters: the log of theta, and the intercept of the zero-inflation model. How can I modify the expression to incorporate my weights? I cannot upload the data as they are large. Any solution using QGIS, the raster package of R, or ArcGIS will be helpful.
أكثر...
Exp(\beta_0+\sum(\beta_i * x_i))/Exp(\beta_0)
In the current case I have two additional parameters: the log of theta, and the intercept of the zero-inflation model. How can I modify the expression to incorporate my weights? I cannot upload the data as they are large. Any solution using QGIS, the raster package of R, or ArcGIS will be helpful.
أكثر...