URL

http://swocker-ssapra.rhcloud.com

Project Description

The goal of this project was to map twitter sentiments to a company to actual Twitter stock prices. To accomplish this, we wrote a Flask-based web application with a SQL database. We used Twitter's API and Yahoo's stock API. Our database has a Company and a Tweet model with a "has_many" relationship (i.e a Company has many Tweets).

Advantages of using a database include really fast indexing of data, and also the potential to store a lot more than a week's worth of historical Twitter data.

Screenshots

Dependencies:

Install command line tools:

xcode-select –install

Install brew:

ruby -e “$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)”

brew doctor

Install Python (2.7):

brew install python

Install virtualenv:

pip install virtualenv

Install mysql (5.5)

Downloading the Code

git clone https://github.com/sameetandpotatoes/Swocker.git

Make sure you have run bin/activate to activate the virtualenv, and then run the following commands to install all necessary dependencies inside the virtualenv:

pip install -r requirements.txt

###secrets.py is organized like this inside the swocker folder:###

ALCHEMY_CODES = [
    ''
]
TWITTER_CODES = [
    {
        'CONSUMER_KEY': '',
        'CONSUMER_SECRET': '',
        'ACCESS_TOKEN': '',
        'ACCESS_SECRET': ''
    },
]

Create a file secrets.py in the app directory with touch secrets.py, and copy the default format.

Fill in your API keys, we'll collect everyone's when we deploy / store them as environment variables in production eventually

Creating the Database

Note: You only have to do this if there isn't an app.db file. Basically, ignore the rest of the instructions, this is just things that I had to do to get things working

Run python db_create.py to create the database
Run python db_migrate.py to migrate the database
Open up the python shell (On Mac, just type python)
from swocker import company
company.load_objects_into_database()
All companies are now created. You can see them all with:
from swocker.models import *
Company.query.all()
Now, to get some tweets in there:
from swocker import tasks
tasks.store_tweets_in_database()

This won't complete all the way. Wait about a day, then rerun.

RabbitMQ and Celery Set Up:

To get this running on a deployed server, you need RabbitMQ and Celery. If you want to try to simulate this on your computer, you need all three of these commands running at the same time:

rabbitmq-server
celery -A app.tasks.celery worker --loglevel=info
celery -A app.tasks.celery beat

You won't see anything for a day or so (you can change the frequency of the task if you want in tasks.py)

Ctrl + D to exit the shell

Running the App

Then run the app with python app.py. You'll see data instantly for the companies that we have tweets for!