These installation instructions are for a general Linux system but should work with minor changes on other OSes (Windows, MacOS, etc..). It assumes that most software will be installed in $HOME/local.
PyICL relies on the following third-party softwares
Python 2.5 or newer. Development has been done on Python 2.6 but it should work on Python 2.5 and Python 2.7. It should not be difficult to get PyICL working on Python 3 but it may not work out of the box.
A C++ compiler to compile the C++ boost.icl library and the boost.python bindings. All development has been done with GCC 4.4.3 or 4.6.3 but other versions and compilers should work.
The Boost C++ libraries: I have used version 1.49 although newer (and perhaps some older) versions should work. Following the commands given at the Boost website is straightforward, e.g.:
./bootstrap.sh --help
./bootstrap.sh --prefix=$HOME/local
./bjam --with-python --prefix=$HOME/local install release
should install the boost Python library and the necessary headers. Once they are installed, you will need to update your LD_LIBRARY_PATH environment variable, e.g.:
export LD_LIBRARY_PATH=$HOME/local/lib:$LD_LIBRARY_PATH
so that the shared objects can be found at runtime.
If you haven’t already, download the PyICL source distribution, unpack it and change into the top level directory.
This python package uses aksetup for installation, which means that installation should be easy and quick. Try:
./configure.py --help
to examine the possible options. By the way, if a configuration option says several ok, then you may specify several values, separated by commas. Something like the following should work:
./configure.py --boost-inc-dir=$HOME/local/include --boost-lib-dir=$HOME/local/lib
python setup.py build
sudo python setup.py install
Configuration is obtained from files in this order:
/etc/aksetup-defaults.py
$HOME/.aksetup-defaults.py
$PACKAGEDIR/siteconf.py
Once you’ve run configure, you can copy options from your siteconf.py file to one of these files, and you won’t ever have to configure them again manually. In fact, you may pass the options --update-user and --update-global to configure, and it will automatically update these files for you.
This is particularly handy if you want to perform an unattended or automatic installation via easy_install or pip.
To check that PyICL has been successfully installed, change to an empty directory and try running the following command:
python -c "import pyicl"
If you see any errors such as:
Traceback (most recent call last):
File "<string>", line 1, in <module>
ImportError: libboost_python.so.1.49.0: cannot open shared object file: No such file or directory
you probably have not updated your LD_LIBRARY_PATH successfully.