Skip to content
Snippets Groups Projects
Select Git revision
  • master
  • 1.8.0
  • 1.7.16
  • 1.7.14
  • 1.7.11
5 results

readme.md

Blame
  • Code owners
    Assign users and groups as approvers for specific file changes. Learn more.

    Scientific Plots

    Create and save plots in scientific style

    Table of Contents

    Overview

    This python module includes useful methods and definitions for various python projects. The focus lies of on the automatic creation of a set of plots, which are designed to be used in scientific journals, dissertations and presentations. The most important components are the definitions of types compatible for numpy, located in types_.py, and the typing stubs in stubs/. These typing stubs are also distributed in this package.

    Plotting

    The easiest way to implement the plotting features provided by this library, is to use one of the predefined function in scientific_plots.default_plots. Alternatively, any plotting functions can be decorated by using the apply_styles decorator in scientific_plots.plot_settings.

    For example, this could look like this:

    import matplotlib.pyplot as plt
    from scientific_plots.plot_settings import apply_styles
    
    @apply_styles
    def plot_something() -> None:
        """Example function."""
        plt.plot(...)
        ...
        plt.savefig("subfolder/your_plot_name.pdf")

    The script will create a bunch of plots and place them in the given location next to your given path. Thus, it is advisable to create a different subfolder for new plots.

    For three-dimensional plots, it is recommended to set the optional argument three_d of the decorator to true:

    @apply_styles(three_d=True)
    def plot_function():
        ...

    Alternatively, this package provides default plot settings in the submodule default_plots. The provided function apply a default design, which should look good in most situations.

    from scientific_plots.default_plots import plot
    
    plot(x, y, "x_label", "y_label", "subfolder/filename.pdf")

    Besides this simple plot, this library also provides the following default plots: plot_fit: Plot data and a fit of this data.

    def fit_function(x):
        ...
        # fit some data
        ...
        return y
    
    plot_fit(
        x, y, fit_function, "x_label", "y_label",
        "subfolder"/"filename.pdf")`

    two_plots: Plot two curves sharing a single y-axis.

    two_plots(
        x1, y1, "label1",
        x2, y2, "label2",
        "xlabel", "ylabel", "subfolder"/"filename".pdf)

    two_axis_plots: Plot two curves with two y-axis in a single graph.

    two_axis_plots(
        x1, y1, "label1",
        x2, y2, "label2",
        "xlabel", "ylabel1", "ylabel2",
        "subfolder"/"filename".pdf)

    All of those functions have the following command-line arguments:

    • logscale: Plot the data double logarithmic.
    • single_log: Plot the x-axis logarithmic.
    • single_log_y: Plot the y-axis logarithmic.
    • xlim: Set the limits on the x-axis manually.
    • ylim: Set the limits on the y-axis manually.

    Preview-Mode

    It is possible to only create a single plot for every called plot function to save computation time. This 'preview' mode can be called by setting the following global variable in scientific_plots.plot_settings to true:

    import scientifc_plots.plot_settings
    scientifc_plots.plot_settings.PREVIEW = True

    Types

    Additional Vector like types for numpy-arrays are provided in scientifc_plots.types_. These types can be used for static type checking using mypy.

    Typing Stubs

    Addtional typing stubs for scipy, matplotlib and numba are provided and installed by this package. These packages do not provide type hints on their own.