edolab 0.0.4

A command line tool for running experiments with `edo`.

edolab

A command line tool for running experiments with edo.

Installation

edolab is pip-installable:

$ python -m pip install edolab

Usage

To use edolab, you need a Python script for your experiment and a place to send the generated data. An valid script must contain:

• a function fitness that takes an edo.Individual instance to be used as the fitness function
• variable assignments for (at least) the essential arguments in edo.DataOptimiser except for families
• definitions of any custom distribution classes to be used
• a list distributions that will be used to create the families

An example of such a script would be something like this:

""" /path/to/experiment/script.py """

from edo.distributions import Uniform


def fitness(individual):
    """ Return the square of the first element of the dataset. """

    return individual.dataframe.iloc[0, 0] ** 2


class MyUniform(Uniform):
    """ A copy of Uniform for demonstrative purposes. """

    name = "MyUniform"
    param_limits = {"bounds": [-1, 1]}


size = 100
row_limits = [1, 1]
col_limits = [1, 1]
max_iter = 5
best_prop = 0.25
distributions = [MyUniform]

For more details on the parameters of edo, see its documentation at: https://edo.readthedocs.io

Then, to run an experiment with this script do the following:

$ edolab run --root="out" /path/to/experiment/script.py

And to summarise the data (for easy transfer):

$ edolab summarise /path/to/experiment/script.py out

It is highly recommended that you use a virtual environment when using edo in or outside of this command line tool as edo uses pickle to store various objects created in a run that may not be retrievable with a different version of Python.

For further details on the commands, use the --help flag on the run and summarise commands.

Contributing

This tool has been made to be pretty bare and could use some padding out. If you'd like to contribute then make a fork and clone the repository locally:

$ git clone https://github.com/<your-username>/edolab.git

Install the package and, optionally, replicate the conda environment:

$ cd edolab
$ python setup.py develop
$ conda env create -f environment.yml
$ conda activate edolab-dev

Make your changes and write tests to go with them, ensuring they pass:

$ python -m pytest --cov=edolab --cov-fail-under=100 tests

Commit, push to your fork and open a pull request!