A tool to sample raster files using vector inputs. This tool originated from necessity and the best alternative was gdallocationinfo which is barely useful for sampling large numbers of points. Using rasterio integrated sampling functionality is an alternative to this tool and there is still a need to do benchmarks of these two tools in order to figure out whether this tool has any benefits at all!

The simplest way to install rasam is by using conda. Simply activate the environment you want to install the tool into and run:

conda install -c conda-forge rasam

Now, you should be able to run rasam -h and get the help displayed.

Alternatively, you can build the tool manually by cloning the repository and using cmake by running

mkdir build
cd build && cmake ..
make

This will place the tool into /build/src/rasam.

Usage is simple and all options are explained in the tool help. To get started, run

rasam <input_raster.tif> <input_points.{gpkg, csv, shp...}> <output_file.{gpkg,csv,shp,...}>

The tool will automatically infer the output file type (if possible) and return all coordinates along with the sampled raster value.