interplot

Create matplotlib and plotly charts with the same few lines of code.

It combines the best of the matplotlib and the plotly worlds through a unified, flat API.

Switch between matplotlib and plotly with the single keyword interactive. All the necessary boilerplate code to translate between the packages is contained in this module.

Currently supported building blocks:

• scatter plots
  • line
  • scatter
  • linescatter
• bar charts bar
• histogram hist
• boxplot boxplot
• heatmap heatmap
• linear regression regression
• line and area fill fill
• horizontal and vertical lines
  • hline
  • vline
• annotations text

Supported

• 2D subplots
• automatic color cycling
• 3 different API modes

  • One line of code

    >>> interplot.line([0,4,6,7], [1,2,4,8])
    [plotly line figure]
    
    >>> interplot.hist(np.random.normal(40, 8, 1000), interactive=False)
    [matplotlib hist figure]
    
    >>> interplot.boxplot(
    ...     [
    ...         np.random.normal(20, 5, 1000),
    ...         np.random.normal(40, 8, 1000),
    ...         np.random.normal(60, 5, 1000),
    ...     ],
    ... )
    [plotly boxplots]
    

  • Decorator to auto-initialize plots to use in your methods

    >>> @interplot.magic_plot
    ... def plot_my_data(fig=None):
    ...     # import and process your data...
    ...     data = np.random.normal(2, 3, 1000)
    ...     # draw with the fig instance obtained from the decorator function
    ...     fig.add_line(data, label="my data")
    ...     fig.add_fill((0, 999), (-1, -1), (5, 5), label="sigma")
    >>> plot_my_data(title="My Recording")
    [plotly figure "My Recording"]
    
    >>> @interplot.magic_plot_preset(interactive=False, title="Preset Title")
    >>> def plot_my_data_preconfigured(fig=None):
    ...     # import and process your data...
    ...     data = np.random.normal(2, 3, 1000)
    ...     # draw with the fig instance obtained from the decorator function
    ...     fig.add_line(data, label="my data")
    ...     fig.add_fill((0, 999), (-1, -1), (5, 5), label="sigma")
    >>> plot_my_data_preconfigured()
    [matplotlib figure "Preset Title"]
    

  • The interplot.Plot class for full control

    >>> fig = interplot.Plot(
    ...     interactive=True,
    ...     title="Everything Under Control",
    ...     fig_size=(800, 500),
    ...     rows=1,
    ...     cols=2,
    ...     shared_yaxes=True,
    ...     # ...
    ... )
    >>> fig.add_hist(np.random.normal(1, 0.5, 1000), row=0, col=0)
    >>> fig.add_boxplot(
    ...     [
    ...         np.random.normal(20, 5, 1000),
    ...         np.random.normal(40, 8, 1000),
    ...         np.random.normal(60, 5, 1000),
    ...     ],
    ...     row=0,
    ...     col=1,
    ... )
    ... # ...
    >>> fig.post_process()
    >>> fig.show()
    [plotly figure "Everything Under Control"]
    
    >>> fig.save("export/path/file.html")
    saved figure at export/path/file.html
    

Resources

• Documentation: https://interplot.janjo.ch
• Demo Notebooks: https://nbviewer.org/github/janjoch/interplot/tree/main/demo/
• Source Code: https://github.com/janjoch/interplot
• PyPI: https://pypi.org/project/interplot/

Licence

Demo

View on NBViewer:

Try on Binder:

Install

pip install interplot

install development branch

pip install git+https://github.com/janjoch/interplot.git@development

active development installation

1. git clone https://github.com/janjoch/interplot
2. cd interplot
3. pip install -e .

Contribute

Ideas, bug reports/fixes, feature requests and code submissions are very welcome! Please write to janjo@duck.com or directly into a pull request.