BeyondPartyLines

This suite will prompt you at your leisure for your personal voting decisions, logging these and allowing you to compare them with Canadian Members of Parliament. Become a member of the House of Commons!

Install

If you are fortunate enough to have acquired a binary/executable for the program (there might be some at wbrenna.ca), simply launch it. Otherwise, you will need Python (I used Python2.6, your results may vary with other versions) and a few dependencies:

BeautifulSoup
matplotlib
numpy
pyttk (which requires Tk wrappers for Python)

The easiest method of installation on systems with a good package manager (for example, apt). Here, simply install the prerequisites:

sudo apt-get install python-numpy python-matplotlib python-tk python-beautifulsoup

Then, run the application with

python beyondpartylines.py

If, for some reason, you cannot install the dependencies as above, this application can be setup in two other ways (see below for additional install notes).

Run

The application window will pop up. Initially the list of votes will be unpopulated. Click "Download" to download votes from the last vote in currentsitting.dat. The window should be populated with votes from the House of Commons, last to first.

More help is provided from the Help menu in program.

Hinton Plots

The Hinton plot deserves a little explaining. This plot allows for the visualization of your votes versus another MPs. Currently the axes are not labelled but the votes begin from the top left (the highest number of bill-vote you have voted Yes, No, or Abstain on) and travel left-to-right, top-to-bottom, in the order of decreasing vote number. The reason labelling is nontrivial is because you may have skipped votes, and vote data may go between parliamentary sittings. For example, if you voted on issues 445, 440, 441, and 404, the Hinton plot would be a 2x2 grid as follows:

445	441
440	404

A large white box means you voted the same way as the MP (Yes-Yes or No-No). A black large box means you directly opposed one another (Yes-No). The small white box means you abstained and the MP voted, while a small black box means you voted and the MP abstained. An empty grey square means you both abstained, or, since the plot must be square, it is filler for the remainder of the plot space. At the top, the correlation index is specified, to say how often you agreed with the MP.

Installation Details

Install with pip

First, ensure you have pip and virtualenv installed. Start a virtualenv with

virtualenv --no-site-packages bpl_env
. bpl_env/bin/activate

Then, you can install the packages required with

pip install -r pipdeps.txt

If you get an error like this:

BUILDING MATPLOTLIB

        matplotlib: 1.1.0
            python: 2.6.6 (r266:84292, Dec 26 2010, 22:31:48)  [GCC 4.4.5]
          platform: linux2

REQUIRED DEPENDENCIES

             numpy: no
                    * You must install numpy 1.1 or later to build
                    * matplotlib.

you likely have to install numpy first. This is a known bug in pip, and it also occurs with setuptools (below), meaning you have to satisfy the numpy dependencies before the installation will go on. You simply need to run

pip install numpy

followed by

pip install -r pipdeps.txt

Then, while you are still in the virtualenv, you can run the software with

python beyondpartylines.py

To exit the virtualenv, type

deactivate

and to reenter the virtualenv (to run the software again) run

. bpl_env/bin/activate

again. It is that easy!

Install with setuptools

This option will automatically install setuptools for you, if you do not have it. You will likely have to preinstall numpy (see above).

sudo python setup.py install

I have never actually been able to get this approach to work, since building numpy and matplotlib with setuptools has always broken on my systems. One thing that may help would be installing the Python headers (python2.6-dev on my system).

Lingering Errors

If you still have problems, feel free to send me an email. One issue that appears common on Ubuntu is the error

No module named _tkagg

or similar. This can be fixed by removing the matplotlibrc file or by changing the backend specified in matplotlibrc.

Also, when the parliament breaks XML links for votes, you might encounter some errors when trying to download votes. Manually edit the file currentsitting.dat with the current session, sitting, and vote number.

If you ever need to correct any votes, you can always roll the vote number back in currentsitting.dat and delete the relevant lines in votes.dat, then re-download and re-vote. Duplicate votes in votes.dat are automatically removed. Remember that MPs are not allowed to roll back votes, though, so perhaps you should follow their example!