TwitterHealth

A webapp to gather 'sickness' related tweet data.

##Features##

• Crowdsource the classification of tweets by allowing users to read tweets and choose if a tweet's content is related to being sick or not.
• Display a visualization of sick tweets across the US
  • show the choropleth by density (# of sick tweets / population)
  • show the choropleth as compared to the average (compare the current amount of tweets in 15 minutes with the average)
• Display a visualization of US/state tweets over the last week.
• mobile/responsive, put it online

##Guide## To Install:

• install node.js
• download this repo (git clone or download the zip file)
• cd to the folder and run npm install
• create an .env file (add with your twitter/api credentials)
• use seperate twitter api keys for production and development if you don't want to get errors/rate limited.
• you can use firebase (100mb free) to store the classified tweets/classifier.
• rediscloud for production (25mb free)

//.env
CONSUMER_KEY=
CONSUMER_SECRET=
ACCESS_TOKEN=
ACCESS_TOKEN_SECRET=
DEV_CONSUMER_KEY=
DEV_CONSUMER_SECRET=
DEV_ACCESS_TOKEN=
DEV_ACCESS_TOKEN_SECRET=
FIREBASE_SECRET=
REDISCLOUD_URL=

To Run:

• run node server
• go to localhost:5000 in your browser

To Build:

• run 'gulp'

To deploy on heroku:

• add the free rediscloud addon
• add heroku config:set NODE_ENV=production
• add all environment variables
  • example: use heroku config:set CONSUMER_SECRET=KEYHERE
• to use gulp on heroku
  • add environment variable: BUILDPACK_URL=https://github.com/timdp/heroku-buildpack-nodejs-gulp.git
• run git push heroku (assuming you added the repo link)