DESCRIPTION

This program calculates a Dempster-Shafer Theory (DST) predictive model. If you have need for a flexible spatial predictive modelling framework but do not know what DST is, please refer to the manual page of dst.combine for an introduction.

Simply specify your raster evidence maps, sample points and a prefix for your result maps and a basic DST predictive model will be calculated. For this, dst.predict will invoke the commands r.dst.bpa, dst.create, dst.source, dst.update and dst.combine
automatically with appropriate arguments.

You can choose to calculate further DST metrics by using the value= option. Refer to the mannual page of dst.combine for their meanings.

You can also specify one or more maps that encode bias for the "NO SITE" hypothesis. See the manual page of r.dst.bpa for an explanation of how this works.

You can also specify attributes that encode bias for the "SITE" hypothesis. Again, see the manual page of r.dst.bpa for an explanation of how this works. (Note: The sbias= option is currently broken. Do not use it!)

If you wish, the program will create an ASCII log file with the combined log output of all invoked commands.

You can specify the -r flage to remove all output maps pertaining to the NULL hypothesis. These are usually only of mathematical interest and can safely be deleted.

If you want to use this frontend for exploratory modelling, you might find it more convenient to dump model parameters to a shell script which can be used to re-run the model at any time and adjust parameters more efficiently. Use the script= option to specify a filename for the script.

Notes

This program was developed as part of the GRASS 5 DST Predictive Modelling Toolkit.
The sbias= option is currently broken. Do not use it!

SEE ALSO

dst.update
dst.create
dst.source
dst.update
r.dst.bpa

AUTHOR

Benjamin Ducke,
University of Kiel, Germany
Last changed: 2005/07/21