DESCRIPTION

r.rand.weight generates a binary raster layer with a random selection of raster cells which are assigned 1. The other cells are assigned NULL (or optionally 0). The change for a raster cell to get assigned a 1 (to get selected) depends on the weight (value) of that cell in the input weight layer.

By default the script is run setting a random seed every time. To ensure that your results are reproducible you can set the seed value under the 'Sample options' tab. See the 'Random number generator initialization' in the r.mapcalc helpfile for more details.

You can play with the probability for a cell to be selected by changing the minimum and/or maximum weights. The script will give a warning if the user defined minimum > minimum raster value or if the user defined maximum is smaller then the maximum raster value. The script will still run as the user may have set this values intentionally.

You can also set the total number of sample points to be selected using the under the 'Sample options' tab. This can be done using an absolute number or as percentage (see the help file of the r.random function for more details).

Examples

See here for examples

See also

r.random, r.random.cells,

AUTHOR

Paulo van Breugel, paulo at ecodiv.org Last changed: $Date$