I'm using the "neuroim" package from CRAN and the function connComp3D() for identifying connected extreme events in climate impacts (the mask is here where vegetation productivity is below a certain threshold). Connected "events" identified by connComp3D() include connections of diagonally neighbouring voxels in a rectangular grid (in 2D, one gridcell is surrounded by 8 neighbours). It would be great to have an option to set this to ignore diagonal connections, i.e., considering that in 2D, each gridcell is only surrounded by 4 neighbours.
This would be a straight-forward way to reduce the size of events (desirable e.g. for my application).
Thanks and congratulations to this useful algorithm!
I'm using the "neuroim" package from CRAN and the function
connComp3D()
for identifying connected extreme events in climate impacts (the mask is here where vegetation productivity is below a certain threshold). Connected "events" identified byconnComp3D()
include connections of diagonally neighbouring voxels in a rectangular grid (in 2D, one gridcell is surrounded by 8 neighbours). It would be great to have an option to set this to ignore diagonal connections, i.e., considering that in 2D, each gridcell is only surrounded by 4 neighbours. This would be a straight-forward way to reduce the size of events (desirable e.g. for my application). Thanks and congratulations to this useful algorithm!