OkStereotypeMe! ==============

Source code for okstereotype.me

• scikit-learn https://github.com/scikit-learn/scikit-learn
• django https://www.djangoproject.com/
• rabbitmq http://www.rabbitmq.com/
• pandas http://pandas.pydata.org/
• Amueller's word cloud https://github.com/amueller/word_cloud
• XKCD plot for matplotlib http://jakevdp.github.com/blog/2012/10/07/xkcd-style-plots-in-matplotlib/

First scrape data from okcupid and store in a pandas dataframe as described in pick_best_model_v1.py and test how different models perform on your dataset.

Second, pickle the data using pick_best_model_v2.py

Third, run the django server using manage.py and run the rpc server using queue/rpc_server.py

• pick_best_model_v1.py should only be used for testing purposes
• multi_q.py is not a good way to create a multicore vectorization of new essays. Use rpc_server.py instead.