virtualenv ========== If you want to merely test out the library, you can do that easily in a *virtual python environment* created with `virtualenv`. **NB:*** You need a build environment for your platform set up as described in the **Installation** section. Setup ----- The following steps set up an evironment: * create a directory and initialize the environment .. code-block:: bash $ mkdir pwwtest $ virtualenv pwwtest * install the required packages .. code-block:: bash $ pwwtest/bin/pip install numpy $ pwwtest/bin/pip install PIL $ pwwtest/bin/pip install matplotlib $ pwwtest/bin/pip install pygraphviz $ pwwtest/bin/pip install javabridge * as an alternative, you can also use all currently installed packages on your host system in the virtual environment. For this, you have to have to all the required packages installed on your host already (see **Installation** section). You can create the virtual environment as follows: .. code-block:: bash $ virtualenv --system-site-packages pwwtest Troubleshooting --------------- * You may need to install the header files for the following libraries for the compilations to succeed: * freetype * graphviz * png * Before you can install `matplotlib`, you may have to upgrade your `distribute` library as follows: .. code-block:: bash $ bin/easy_install -U distribute * On Ubuntu, follow `this post `__ to install all the required dependencies for PIL: .. code-block:: bash $ sudo apt-get build-dep python-imaging * To enable support for PIL on Ubuntu, see `this post `__: .. code-block:: bash $ sudo ln -s /usr/lib/`uname -i`-linux-gnu/libfreetype.so /usr/lib/ $ sudo ln -s /usr/lib/`uname -i`-linux-gnu/libjpeg.so /usr/lib/ $ sudo ln -s /usr/lib/`uname -i`-linux-gnu/libz.so /usr/lib/ * Based on `this post `_ do the following to install matplotlib with tkinter support: .. code-block:: bash sudo apt install tk-dev /my/virtualenv/dir/bin/pip --no-cache-dir install -U --force-reinstall matplotlib