DESCRIPTION

The t.rast.kappa calculate kappa parameter in a space time raster dataset. It can calculate kappa values using two different algorithm, the one provided by r.kappa and the one provided by SciKit-Learn metrics library

EXAMPLE

Using r.kappa as backend

In this example t.rast.kappa is using r.kappa as backend, it return the results inside the /tmpem> directory into files with mystrds as prefix. weight option is not considered using r.kappa as backend.

        t.rast.kappa -k strds=mystrds output=/tmp/mystrds

Using SciKit-Learn as backend, text as output

In this example t.rast.kappa is using SciKit-Learn metrics library as backend, without output option, the module print the results to standard output

        t.rast.kappa strds=mystrds where="start_time >= '1999-12-01' weight='linear'

Using SciKit-Learn as backend, map as output

In this example t.rast.kappa is using SciKit-Learn metrics library as backend, the output is a map with the kappa values calculated for each pixel. The splittingday option is required to split the space time raster dataset in two groups and analyze them; the two groups must have the same number of maps, otherwise and error will be reported.

        t.rast.kappa -p strds=mystrds output=mykappa splittingday='2005-01-01'

AUTHOR