PyVO is a package providing access to remote data and services of the Virtual observatory (VO) using Python.
Its development was launched by NSF/NASA-funded Virtual Astronomical Observatory (VAO, www.usvao.org) project (formerly under the name VAOpy) as part of its initiative to bring VO capabilities to desktop. Its goal is allow astronomers and tool developers to access data and services from remote archives and other web resources. It takes advantage VO standards to give access to thousands of catalogs, data archives, information services, and analysis tools. It also takes advantage of the general capabilities of Astopy (and numpy), and so a secondary goal is to provide a development platform for migrating more VO capabilities into Astropy.
Information about this package, including latest releases, can be found at:
Source code can be found on GitHub at:
This implementation requires the following prerequisite packages:
- numpy (1.6.0 or later)
- astropy (0.2 or later)
These must be installed before install PyVO.
As an Astropy affiliate, this package uses the Astropy build infrastructure.
Releases of PyVO are available from PyPI <https://pypi.python.org/pypi/pyvo>; thus, it and its prerequisites can be most easily installed using pip:
pip install pyvo
Alternatively, you can download and unpack a source tar-ball (pyvo-x.x.tar.gz). To install directly into the python installation, type as root user inside the distributions root directory:
python setup.py install
To install into a special directory called, say, $MYPYVO (which need not require root permission), first be sure that astropy and numpy are in your PYTHONPATH (if they are also installed in a non-standard place). Next, try:
python setup.py install --home=$MYPYVO
To just try out PyVO in this directory, you can build it in "developer" mode via:
python setup.py build_ext --inplace
In this mode, update your PYTHONPATH to include the directory containing this file.
The scripts directory contains instructive examples of the use of pyvo.
The PyVO unit tests have not yet been integrated into the py.test framework used by Astropy. (They currently use the standard unittest module.) Unit tests are contained in the tests subdirectory. Before running a test, be sure that the directory containing the tests subdirectory (.) is in your PYTHONPATH (along with astropy and numpy). Then, to run a particular unit test script, type:
python tests/path-to-script.py
For example:
python tests/dal/testSIA.py
If the script contains the string "NeedsNet" in its name, it requirs access to the network to work properly. All other files do not require the network, working off of local input files our local services executed locally on-the-fly.
A few files aggregate the running of tests from several individual unit test files. In particular:
- testNoNet.py -- runs all unit tests that do not require access to
the network to work.
- testNeedsNet.py -- all tests that require the network to work in
order to complete.