contour-maps-visualisation

Objective

This library offers a way to visualize multiple 2D-distribution where each distribution has one categorical variable. It tries in particular to help the viewer to achieve a better understanding of the distributions. It ships with three types of visualizations:

where each can be used with the others as pleased. There are four kinds of options for the visualization in total. Contours, contour-lines, crosses and pie-charts.

Prerequisites

Install submodules with:

git submodule update --init

Make sure python 3.5 or higher and pip3 are installed. Then install the package with:

pip3 install .

Note

For Development use:

pip3 install -e .

or if the package is used internal its also possible to install the dependencies with:

pip3 install -r requirements.txt

Jupyter Interactive Notebook should be installed with requirements. If not install it manually

Quickstart

import the Modules:

from contour_visualization import example_data, picture_plot

create a list of distributions with:

_, _, gaussians, _ = example_data.generate_four_moving_gaussians(size=100)

and get your first visualisation with:

picture_plot.plot_images(gaussians, contours=True, contour_lines=True, contour_line_level = 2)