A real-time 3D visualization of Tweets from around the world.

This web app attaches to the Twitter API stream/filter and runs rudimentry sentiment analysis on Tweets with geo data. Tweets are published via PubNub websockets and plotted to a 3D globe.

Inspired by the Web GL Globe Chrome Experiment and the PubNub Real-Time WebGL Visualization.

 Video Capture | Running Demo (Availability of this server may fluctuate)

Install Node.js.

Clone GitHub repo:

https://github.com/twitterdev/twitter-stream-globe.git

Create a Twitter app and PubNub account:

• Create a Twitter application.
• Create a PubNub account (it's free).

Create a config.json file using config.sample.json as a template. Fill in your Twitter App API and PubNub keys.

Install node module dependencies:

npm install 

Run application:

npm start

Go to http://localhost:3000 in your browser.

This application is ready to run on a free OpenShift or Heroku account, and can be deployed with Git.

You can deploy to OpenShift with rhc, by adding your own keys to the following command:

rhc app create twglobe nodejs-0.10 \
  --from-code=http://github.com/twitterdev/twitter-stream-globe.git \
  NODE_ENV=production \ 
  TWITTER_CONSUMER_KEY=YOUR_TWITTER_CONSUMER_KEY \
  TWITTER_CONSUMER_SECRET=YOUR_TWITTER_CONSUMER_SECRET \
  TWITTER_ACCESS_TOKEN=YOUR_TWITTER_ACCESS_TOKEN \
  TWITTER_TOKEN_SECRET=YOUR_TWITTER_TOKEN_SECRET \
  PUBNUB_PUBLISH_KEY=YOUR_PUBNUB_PUBLISH_KEY \
  PUBNUB_SUBSCRIBE_KEY=YOUR_PUBNUB_SUBSCRIBE_KEY

Before deploying to Heroku, set your environment config vars to mirror config.json, and set NODE_ENV to "production."

• Twitter API statuses/filter stream
• Twitter REST API Rate Limiting
• AngularJS
• PubNub AngularJS SDK
• Three.js