GRASS logo


r.random - Creates randomly placed raster cells or vector points
Creates a raster map and vector point map containing randomly located cells and points.


raster, sampling, vector, random, level1


r.random --help
r.random [-snzb] input=name [cover=name] npoints=number[%] [raster=name] [vector=name] [seed=integer] [--overwrite] [--help] [--verbose] [--quiet] [--ui]


Generate random seed (result is non-deterministic)
Generate points also for NULL cells
Generate vector points as 3D points
Input raster values will be used for Z coordinates
Do not build topology
Do not build topology for vector points
Allow output files to overwrite existing files
Print usage summary
Verbose module output
Quiet module output
Force launching GUI dialog


input=name [required]
Name of input raster map
Name of cover raster map
npoints=number[%] [required]
The number of points (or cells) to generate
The number of vector points or raster cells to generate, possibly as a percentage of number of cells
Name for output raster map
Name for output vector map
Seed for rand() function

Table of contents


The module r.random creates a raster map with values in random places. Alternatively, it creates random vector points at these places. Number of random cells or points can be a fixed number or a percentage of cells from the input. By default, generated cells or points will be subset of non-NULL cells of the input. Resulting raster map consists of original cell values at the selected random locations and NULL (no data) values elsewhere.

Placement of cells and points

The module allows the user to create a raster map and/or a vector points map containing coordinates of points whose locations have been randomly determined. The module places these randomly generated vector points within the current computational region and raster mask (if any), on non-NULL raster cells in a user-specified raster map. If the user sets the -n flag, points will be randomly generated across all cells (even those with NULL values). Cells in the resulting raster overlap with the cells of the input raster based on the current computational region. Points in the resulting vector map are placed in cell centers of these cells.

Number of cells and points

The user may specify the quantity of random locations to be generated either as a positive integer (e.g., 10), or as a percentage of the raster map's cells (e.g., 10%, or 3.05%). The number of cells considered for the percentage reflects whether or not the -n flag was given. Options are 0-100; fractions of percent may be stated as decimals (e.g., 66.67%, or 0.05%).


The cell values and corresponding category names (if present) associated with the random point locations in the input map are assigned to the newly generated cells in the raster map. If the -n is specified, then a unique entry is made for the value used where the input was NULL. This value is at least 1 less than the smallest value in the input raster and is given a medium gray color.

If a cover raster map is specified, values are taken from the cover raster map instead of the input raster map. If a cover raster map is specified and the cover map contains NULL (no data) values, these points are suppressed in the resulting vector or raster map.

Vector output

The vector file created by r.random contains vector points that represent the center points of the randomly generated cells. A value attribute contains the cell value of the input raster (or the assigned value when -n is used). If a cover map is additionally specified, a second column covervalue is populated with raster values from the cover map.

If the user sets the -b flag, vector points are written without topology to minimize the required resources. This is suitable input to and other vector modules.


To decide on the number of points r.random will create, use r.univar, g.region, or r.univar is the fastest way to obtain number of non-NULL cells and NULL cells in a raster map given the current computational region and raster mask:
r.univar map=inputmap
The text output contains total number of null and non-null cells (called cells in the machine-readable shell script style output), total null cells (null_cells), and number of non-null cells (n). Alternatively, you can use the following to examine the computational region and the raster map:
g.region -p map=inputmap units=c null="*" nsteps=1

To create random vector point locations within some, but not all, categories of a integer input raster map (aka CELL raster map), the user must first create a reclassified raster map of the original raster map (e.g., using the GRASS module r.reclass) that contains only the desired categories, and then use the reclassed raster map as input to r.random.


Random 2D vector elevation points

Random vector elevation points sampled from elevation map in the North Carolina sample dataset region, result stored in 2D vector map:
g.region raster=elevation -p
r.random elevation vector=elevrand n=100 elevrand
v.univar elevrand col=value type=point

Random 3D vector elevation points

Random vector elevation points sampled from elevation map in the North Carolina sample dataset region with collocated values sampled from landuse map, result stored in 3D vector map:
g.region raster=elevation -p
r.random -z elevation cover=landclass96 vector=luserand3d n=100

# data output (value: elevation, covervalue: landuse class): luserand3d



Dr. James Hinthorne, GIS Laboratory, Central Washington University

Modified for GRASS 5.0 by Eric G. Miller

Cover map support by Markus Neteler, 2007


Available at: r.random source code (history)

Latest change: Thu Feb 3 11:10:06 2022 in commit: 73413160a81ed43e7a5ca0dc16f0b56e450e9fef

Main index | Raster index | Topics index | Keywords index | Graphical index | Full index

© 2003-2022 GRASS Development Team, GRASS GIS 8.0.3dev Reference Manual