These materials are available as a resource for a workshop presented by Fast Forward Labs.
If you are attending the workshop and would like to run the code on your own machine as we go through (which is not necessary, but will increase your understanding), then there are some things you should do before the workshop.
Download this repository. To do this use
git cloneif you have git on your machine, or click the green "Clone and Download" button, then "Download ZIP".
Ensure you have the requirements installed. You'll need:
- Python 2.7 or higher
- jupyter notebook, numpy and pandas
- scikit-learn 0.18.2 or higher (the LDA algorithm was added in 0.18.1 but had a bug that was fixed in 0.18.2)
- Optionally, if you would like to build an interactive visualization of a topic model then you'll need pyLDAvis 1.5 or higher, but this is not absolutely necessary to run the core of the notebook.
These dependencies can be installed using
conda(if you have Anaconda Python) or
pip install(if you are using standard tools, in which case do
pip install -r requirements.txt).
Alternatively, if you just want to follow along with the presentation without running the code locally, view lda_rendered.ipynb on Github by clicking this link in your browser.