keen-congress-vote

Experiment using the Keen NodeJS api to import and then use the Keen workbench to analyze congressional vote data

I've been trying out https://keen.io/ for some analytics and decided to try a little more challenging data. This time I am using congressional voting data that can be found off of https://www.govtrack.us/congress/votes

The GovTrack data has a number of different fields that can be used including the following

person: bioguideid, birthday, cspanid, firstname, gender, gender_label, id, lastname, link, middlename, name, namemod, nickname, osid, pvsid, sortname, twitterid, youtubeid

person_role: caucus, congress_numbers, current, description, district, enddate, id, leadership_title, party, person, phone, role_type, role_type_label, senator_class, senator_rank, startdate, state, title, title_long, website

vote: category, category_label, chamber, chamber_label, congress, created, id, link, missing_data, number, question, question_details, related_amendment, related_bill, required, result, session, source, source_label, total_minus, total_other, total_plus, vote_type

voter_type
voter_type_label

Note that Keen IO pricing has a free tier that includes up to 50,000 events per month. If you're using Keen to store data for an open data project you can request a higher limit. Just send them an email.

To install and use this you can do the following.

Find a vote data set (JSON) that you want to process (such as https://www.govtrack.us/api/v2/vote_voter?vote=116114).

• install NodeJS
• create your Keen account and a project
• export environment variables for your project info (these will be used by the code)
  • export projectId=xxx
  • export writeKey=xxx
  • export readKey=xxx
  • export masterKey=xxx
• git clone https://github.com/dougtoppin/keen-congress-vote.git
• npm install keen.io
• node keen-congress-vote/congress-vote.js 116114

The data should have been uploaded to your Keen project as a new event collection with the name congress-vote-voteId.

The data set consists of an array of records, one for each person.

Next use the Keen workbench to take a look at your data.

• go to the Keen workbench for your project (called congress-vote)
• select the Event Collection called congress-vote with the voteId suffix that you specified as the argument
• select Analysis Type called count_unique
• select Target Property called person_role.party
• select Run Query

You should see the result of the query as 2 indicating that 2 values for party were found across the data set.

Another query you could try would be to find each of the unique birthdays from the list of members.

• pick select_unique for the analysis type
• select person.birthday for Target Property
• select Run Query

The output should look something like

Numerous variations of queries can be run with different selections for analysis type and target property. Note that you can consume the output from the query into other applications by requesting the data using the Query Url.