DataPlot

A simple Python class wrapping separate data preparation and plot building steps in a matplotlib pipeline.

Separating the data preparation and plot building:

• makes it easy to provide the exact data in a plot.
• makes the data in a plot available for tests etc., useful when extracting the data requires non-trivial processing and other questions about the data arise.

The basic idea is a that a JSON encodable dict of data is generated by the data preparation function, and this dict is used by the plot building function to make the plot. The DataPlot class is just a thin wrapper around this idea which handles saving the plots in different formats and sets up a suggested skeleton for the dict:

    {
        # author info. etc.
        '_metadata': {
            'generated': time.asctime(),
            # a pointer to this class, to explain JSON format
            'relation': "https://github.com/tbnorth/dataplot",
        },
        # general plot data, title, x label, etc.
        'ax': {'x': {}, 'y': {}},
        'ds': {},  # x, y, etc., lists of the same length
        'kv': {},  # arbitrary key / value info.
    }