Skip to content

pysiss/pysiss

Repository files navigation

Introduction to pysiss

We want to make it easy to perform geological data analysis based on SISS services (including boreholes & geological measurements, raster and vector map data, and vocabularies) within Python using your favourite Python libraries.

This library is released under the CSIRO BSD/MIT license, whose terms are available in the LICENSE.md file.

Warning - this library is in a pre-alpha state and could change without warning.

Installing via pip

The easiest way to get pysiss is to install the version hosted on PyPI via pip.

You will have to have the standard numpy/scipy/matplotlib stack. You also need to have the GDAL libraries and Python bindings installed.

For more details on installing these libraries for Windows, Mac and Linux please see INSTALL.md

Once the dependencies are installed pysiss and all its Python dependencies should be installable using pip

pip install pysiss

This should pull all the dependencies for pysiss automatically. Depending on where you're trying to install the library, you may need administrator priviledges to install the library. If you don't have adminstrator rights for your system, then you can install it locally using

pip install pysiss --user

which will install it under your home directory (usually somewhere like ~/.local/lib/).

Building and installing pysiss from source

pysiss has quite a few dependencies, most of which come from the numpy/scipy/matplotlib ecosystem. You also need to have the GDAL libraries and Python bindings installed. See the instructions in INSTALL.md for getting access to this stack.

Once you've got the numpy/scipy/matplotlib stack plus the GDAL libraries installed, you need:

  • pandas for data munging,
  • shapely, which lets you deal with vector GIS data nicely
  • owslib for calls to SISS services, and
  • simplejson and lxml for dealing with JSON, XML and text data for some of the queries.

If you want to run the examples, you might also want to consider

  • ipython to run the iPython notebook examples
  • folium a wrapper for leaflet.js maps
  • mpld3 a matplotlib to D3 wrapper

These are optional but pretty kick-arse libraries which you should install and play with anyway.

If you've installed all the libraries above, all you should need to do is enter the python directory, and execute

python setup.py install

or

python setup.py install --user

which should build and install pysiss for your system.

Unit tests

To run the unit tests, just go to the base directory and execute

python setup.py test

They only take a couple of seconds to run and should all pass unless I've screwed something up... The unit tests use the vanilla unittest framework, so should play nicely with your favourite testing framework should you prefer to use something else.

Building documentation

If you want to build your own version then you will need to have a version of sphinx installed -- you can check by doing the following at a terminal prompt:

python -c 'import sphinx'

If that fails grab the latest version of and install it with::

pip install Sphinx --upgrade

Now you are ready to build your docs, using make (or run the batch script make.bst if you're on Windows):

cd docs && make html

(or latexpdf if you want a LaTeX versionm, or epub for ePub format - type make to see all the options). The documentation will be dumped under build/. For HTML, if you point a browser to docs/build/html/index.html, you should see a basic sphinx site with the documentation for pysiss. For LaTeX you can open docs/build/latex/pysiss.pdf in your favourite PDF viewer to browse the documentation.

Contributing

For a list of contributors, see AUTHORS.md.

We'd love to have more people use the library and contribute to it. If you've pulled this from the public repository on CSIRO install of Stash (stash.csiro.au), then you might like to check out the mirrored repository on Github or Bitbucket which should make it easier for non-CSIRO types to fork and hack away.

We like unit tests and documentation - feel free to contribute your own.

For more details, feel free to contact Jess: his email is jesse.robertson with CSIRO's domain (google it).