Closed jlemond closed 2 years ago
The Krige
class doesn't support external drift kriging at the moment.
I extended the original class last year to be able to use universal kriging, but since external drift kriging needs a specified drift at the output locations, I didn't know how to properly pass them.
This was decided here: https://github.com/GeoStat-Framework/PyKrige/pull/158#issuecomment-659968767
We need further discussion about that internally.
For now you could use ext_drift_grid
of the Krige
class And use a grid defined with the external drift, where all values are extracted during calculation. But that only works for 2D.
Hello,
I currently use your very useful kriging package PyKrige for Universal Kriging with 2 external drifts (gridded temperature and gridded elevation) as follow:
where: lon= longitudes of stations, lat= latitudes of stations, temp= the temperature values of stations that I want to krige, z= the elevation of stations used as external drift, raw= the temperature values used as external drift, gridx= vector of longitudes, gridy= vector of latitudes , dem_indo= gridded elevation data , raw_temp_fit.values= gridding temperature data
I would like to use Kriging CV to search optimal parameters amongst different variograms and number of bins (nlags).
I’ve tried following code:
As I don’t find the specified drift argument in the Krige() function, I wonder how to manage the drift values. Could you help me?
Thanks in advance, Julien