- Allow for specification of list-of-lists vs. dictionary-of-keys or other sparse data structures
- Write C&F's prime-based one-sparse estimator function to enable negative weight updates
- Write setup.py, init.py
- Write docstrings for all functions within samplers.py, esp. for the L0-Sampler
- Include installation instructions
- Include license explicitly
- Write a "Droplet is a library for sampling and sketching massive data streams in Python" thing
Credits to: https://hacks.mozilla.org/2013/05/how-to-spread-the-word-about-your-code/ Also: http://blog.clojurewerkz.org/blog/2013/04/20/how-to-make-your-open-source-project-really-awesome/