evoeuler 1.0.0

EvoEuler: Evolving Euler and Venn Diagrams with Specific Properties

EvoEuler: Evolving Euler and Venn Diagrams with Specific Properties

EvoEuler searches for Euler and Venn diagrams with a simple closed shape using a multi-objective evolutionary algorithm. Both a web interface and an application programming interface (API) are available.

EvoEuler allows users to:

• specify optional constraints (symmetric diagram and congruent shape) before optimization.
• obtain a set of non-dominated solutions (diagrams), each evaluated with five objectives (missing zone, unwanted zone, triple point, concurrency, and disconnected zone) that quantify the violation of well-matchedness or well-formedness properties.
• explore trade-offs among non-dominated solutions via parallel coordinates plots that visualize the objective values of all solutions.

Installation

Run the following command to install EvoEuler:

pip install evoeuler

Web Interface

Run the following command to launch the web interface:

evoeuler

Compared to the API, the web interface additionally allows users to draw, redraw, round, smooth, or view custom simple closed shapes, which can then be used for diagram generation.

API

The example below shows the use of all functions provided by the API.

The supported algorithms for the API are: NSGA-II and NSGA-III.

The supported shapes for the API are: circle, ellipse, rectangle, rounded_rectangle, equilateral_triangle, rounded_equilateral_triangle, square, rounded_square, pentagon, rounded_pentagon, hexagon, rounded_hexagon, heptagon, rounded_heptagon, octagon, rounded_octagon, nonagon, rounded_nonagon, decagon, and rounded_decagon.

Note that if symmetric is set to True, congruent will be automatically set to True.

"""Search for symmetric Venn-3 diagrams with congruent ellipses."""

import matplotlib.pyplot as plt

from evoeuler import diagram, obj_values, optimize, pcp, shape_params, zone_info

if __name__ == "__main__":
    # Start an optimization.
    result = optimize(
        zones={"A", "B", "C", "A B", "A C", "B C", "A B C"},
        shape="ellipse",
        algorithm="NSGA-II",
        n_gen=100,
        n_process=1,
        time=None,
        symmetric=True,
        congruent=True,
    )

    # Print the number of non-dominated solutions.
    print(f"Number of non-dominated solutions: {len(result)}")

    # Show the parallel coordinates plot that visualizes the objective values
    # of all non-dominated solutions.
    fig = pcp(result)
    fig.show()

    solution_idx = 0  # Solution index

    # Show the diagram of a non-dominated solution.
    diagram(
        result[solution_idx], show_zone_labels=False, shade_unwanted_zones=True
    )
    plt.show()

    # Print the objective values of a non-dominated solution.
    print("Objective values:")
    print(obj_values(result[solution_idx]))

    # Print the zone information of a non-dominated solution.
    print("Zone info:")
    print(zone_info(result[solution_idx]))

    # Print the shape parameters of a non-dominated solution.
    print("Shape params:")
    print(shape_params(result[solution_idx]))

Examples

Some example scripts for searching Euler or Venn diagrams are in ./examples.

Contact

If you have any questions, please contact Chi Ho Chan at [e-mail].