Sunday, October 21, 2012

New NOMINATE data!

It's been a long time since I've updated the NOMINATE data stored on this site. In fact, I've never updated it, because the memory of copying responses to roll call votes hundreds of times and pasting them into a spreadsheet turned me off to the idea.

Fortunately, I've been learning some programming (in Ruby and in Python), so I wrote a program to build the voting list for me instead. It's much pleasanter this way! I've also tried to improve the graphs by getting to know R better (as you can tell from above). Now the whole process takes only 5 or 10 minutes.

Within the next week I expect to release the Ruby code for my vote sorting program (I already posted it at Google Drive, but I want to clean it up a bit), along with simple instructions for getting W-NOMINATE data yourself. And hopefully I will be able to get data from earlier years, to add another dimension of fun to this.

Can't wait for the next legislative session to start! It will be much more interesting, examining this data in real time! It'll be possible to use this data to try to predict the outcomes of House and Senate votes in advance, among other things.

--The new data, including more graphs of the House and Senate and a list of legislators, is over here.

PS: William O'Brien didn't have a score before, because he was practically always "Presiding" instead of voting. He now has enough votes to get a score -- (0.72, -0.69).

No comments:

Post a Comment