Skip to content

Latest commit

 

History

History
141 lines (96 loc) · 3.71 KB

README.md

File metadata and controls

141 lines (96 loc) · 3.71 KB
Raw

TFGP

A Python Genetic Programming package for biinary classification and regression that uses tensorflow for faster performance.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

Requires python3

Tensorflow: Any version from 1.4 untill 2.x, before 2.x there is a need to declare eager execution, from 2.x onwards, that is done automaticaly.

Sckit-learn: Any of the latest versions

pip install tensorflow
pip install pandas
pip install sklearn
pip install multiprocessing
pip install pickle

Modules

A brief explanation

Data Processing

DataHandler

This file loads and preprocesses(if defined) the dataset being used.

Genetic Operators

GeneticOperators

This file contains the implementation of both crossover and mutation operators.

Population Operations

Tree
Forest
Selection
ReproductionHandler

Utility

Utils
SavingHandler
excelScript

Utils contains the implementations of protected operations and math fucntions with multiple applications.

SavingHandler contains the functions used to data.

excelScript is a simple script used to merge all csvs in a folder into a single one, uses include joining the individual results from runs for easier visualization and analysis.

Datasets

This folder contains datasets that can be used to test this package.

Breast Cancer Wisconsin
GAMETES
Heart
Ionosphere
Parkinsons
Sonar

How to use

The main file is th GP.py file. This is the file used to run the algorithm as it contains the evolutionary cycle.

Parameter List

Parameters Description
csvname name of the dataset to use
dsetpath path to the dataset folder
forest_type Can be ramped_forest, full_forest or grow_forest
fpath path to the individuals folder
loadname name of the population to load
ngens number of generations
nruns number of runs
popsize population size, the number of individuals
resume False to start a fresh run, True to laod a population and start from there (Future Work)
savename name to save the population from the last complete generation
savesheetdir path to the folder containing the resulting csv files
sheetname name of the reasulting csv file with training and testing values from each generation
tsize standard tournament size
ttype tournament type, 1 is standard, 2 is double

Parameter List (Tree)

Parameters Description
biFunctions list of binary functions that can be used
uniFunctions list of unary functions that can be used
maxDepth maximum depth of the first generated trees (root starts at 0)

Parameter List (Reproduction)

Parameters Description
MUTATIONPERCENT percentage of the mutation operator
CROSSOVERPERCENT percentage of the crossover operator

Contributing

Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests to us.

Versioning

We use SemVer for versioning. For the versions available, see the tags on this repository.

Authors

See also the list of contributors who participated in this project.

License

This project is licensed under the MIT License - see the LICENSE.md file for details.

Acknowledgments

  • Hat tip to anyone whose code was used
  • Inspiration
  • etc