All of the components on which this was originally built have implemented numerous changes since I started with this. While the general approach likely still works, or can be made to work, I'd proceed with caution before using this to generate anything that could get you fired.
- Accept a Pandas DataFrame
- Bin the contained data on a user specified field
- Default to 4 bins, but allow user defined bins to be passed on to pd.cut
- Calculate the % occurance of each bin
- Create meaningful x-axis labels from the Categorical Index
- Allow for explicit labels to be passed to pd.cut
- Return a Bokeh figure with the graphed data
or developed on
- pandas 0.19.2
- numpy 1.12.0
- bokeh 0.12.4