PyX requires Python 3.2 or newer and a TeX installation (including Type1 fonts). Try to start 'python', 'tex' and 'kpsewhich cmr10.pfb' (the latter should issue a full path of the requested Type1 font).
PyX can be installed via pip straightforwardly. It can be run as an zipped egg. As PyX is hosted on PyPI, it can be directly downloaded and installed by pip:
pip install pyx
For debugging, you can call the pyxinfo() method immediately after the import to get some details written to stderr.
PyX can be run without any installation. To test your environment copy the file hello.py from the examples directory to the main pyx directory (where the subdirectory pyx containing the modules is located) and type 'python hello.py'.
The installation of PyX is pretty straightforward if you have installed the Python distutils package.
First, you have to decide which C extension modules you want to build. This can be done by setting the respective flags in the setup.cfg config file. By default no C extension modules are built and appropriate fallbacks will be used instead.
The build_t1code option enables building of an extension module, which enables faster coding/decoding of Type 1 fonts. The only requisites for building this module are the Python header files and a C compiler. Note that the C compiler has to suit the Python distribution you are using.
The second extension module pykpathsea provides Python binding for the kpathsea library, which enables fast searching for files in the TeX/LaTeX directory hierarchy. You will need the header files of this library, which unfortunately are not included in many standard TeX distributions. Note that the fallback, which uses the kpsewhich program, works equally well, although it is not as efficient as using the library directly. If you want to build the C extension module, you may also have to specify the location of the kpathsea header files and of the library itself in the setup.cfg file.
After you have adapted the setup.cfg file to your needs, you can either build the extension modules locally by running
python setup.py build_ext -i
or install PyX system wide by using
python setup.py install
or an appropriate variant thereof.
For development checkout the PyX repo from github:
git clone https://github.com/pyx-project/pyx
To prepare the develpment environment create a virtual env and run
pip install -e .[dev]