Lots of changes.
- Fixed Coveralls, nosetests.
- Changed k-means clustering to use larger values of k in squared steps, e.g. 4, 9, 16, 25.
- `k_range` argument passed to clustering functions is now `k_vals` since it now accepts lists of values rather than just start and end values for a range.
- Changed Pandas data frame sort methods from `df.sort()` to either `df.sort_index()` or `df.sort_values(by='colname')`.
- Updated README.md.