topic_models_using_rosetta

This repository is intended to serve as a tutorial for running topic models on a sample of 5 text documents using the Python packages rosetta.

To follow the tutorial:

1. Clone or download this repository to your local machine.
2. Install the rosetta package (updated for Python 3) from the following forked branch https://github.com/chsuong/rosetta/tree/update. For example, use the following command on your terminal if you have pip installed:

pip install git+https://github.com/chsuong/rosetta.git@update

3. Install python packages os, panda, and random.
4. Install Vowpal Wabbit to be used via the terminal of your local machine. For example, install it using the following command on your terminal if you have brew installed:

brew install vowpal-wabbit

4. Open the IPython Notebook "topic_models_using_rosetta.ipynb" or the Python script "topic_models_using_rosetta.py" in Python 3.
5. Change the file paths in the notebook and the script for your own use.

##Notes:

• data/raw is the directory where the documents are stored as text files.

• This tutorial is based on the tutorial at https://github.com/columbia-applied-data-science/rosetta.

• See https://github.com/columbia-applied-data-science/rosetta for more tutorials and information about the rosetta package.