examples_and_data.rst

Examples and data

The quickstart runs on synthetic terrain and needs no downloads. The example notebooks are different: they work on real rasters and vector files that ship separately. The xrspatial command-line tool fetches both.

The command-line tool

Installing the package puts an xrspatial command on your PATH. Run it inside the same environment where you installed the library. Three subcommands download the starter material:

xrspatial examples         # notebooks and the data they use
xrspatial copy-examples    # notebooks only
xrspatial fetch-data       # data only

Each one writes into a folder named xrspatial-examples in your current directory. A fourth subcommand, xrspatial clean-data, removes downloaded data files.

The command list and per-command options are available with --help:

xrspatial --help
xrspatial examples --help

If xrspatial reports a missing command, install the CLI helper with pip install pyct.

Getting started with the examples

Download everything and open the notebooks:

xrspatial examples
cd xrspatial-examples
jupyter notebook

Notebooks live under user_guide/ and cover surface analysis, proximity, classification, remote sensing, and the rest of the modules. Downloaded datasets land in xrspatial-examples/data/. The same notebooks are rendered in the :doc:`user guide </user_guide/index>` if you would rather read than run.

Re-running a download overwrites nothing by default. Pass --force to replace files that already exist, or --path to write somewhere other than the current directory:

xrspatial examples --path ~/xrspatial-demo --force

Using your own data

You do not need the example data to use the library. Any 2D xr.DataArray works as input, whether you build it yourself, read it with :doc:`open_geotiff </reference/geotiff>`, or get it from another library. The :doc:`quickstart` shows the no-download path end to end, and :doc:`/reference/geotiff` covers reading local files and http(s) / s3 / gs / az URLs.