OPFython: A Python-Inspired Optimum-Path Forest Classifier
Welcome to OPFython.
Note that this implementation relies purely on the standard LibOPF. Therefore, if one uses our package, please also cite the original LibOPF authors.
Have you ever wanted to classify data into labels? If yes, OPFython is for you! This package is an innovative way of dealing with an optimum-path forest classifier. From bottom to top, from samples and datasets to the actual classifier, we will foster all research related to this new trend.
Use OPFython if you need a library or wish to:
- Create your datasets;
- Design or use pre-loaded state-of-art classifiers;
- Mix-and-match different strategies to solve your problem;
- Because it is cool to classify things.
Read the docs at opfython.readthedocs.io.
OPFython is compatible with: Python 3.6+.
Package guidelines
- The very first information you need is in the very next section.
- Installing is also easy if you wish to read the code and bump yourself into, follow along.
- Note that there might be some additional steps in order to use our solutions.
- If there is a problem, please do not hesitate. Call us.
Citation
If you use OPFython to fulfill any of your needs, please cite us:
@misc{rosa2020opfython,
title={OPFython: A Python-Inspired Optimum-Path Forest Classifier},
author={Gustavo Henrique de Rosa and João Paulo Papa and Alexandre Xavier Falcão},
year={2020},
eprint={2001.10420},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
Datasets
In search for datasets? We have some already pre-loaded into OPF file format. Just check them out at our website!
Getting started: 60 seconds with OPFython
First of all. We have examples. Yes, they are commented. Just browse to examples/, chose your subpackage, and follow the example. We have high-level examples for most tasks we could think.
Alternatively, if you wish to learn even more, please take a minute:
OPFython is based on the following structure, and you should pay attention to its tree:
- opfython
- core
- heap
- node
- opf
- subgraph
- math
- distance
- general
- random
- models
- knn_supervised
- semi_supervised
- supervised
- unsupervised
- stream
- loader
- parser
- splitter
- subgraphs
- knn
- utils
- constants
- converter
- decorator
- exception
- logging
Core
Core is the core. Essentially, it is the parent of everything. You should find parent classes defining the basis of our structure. They should provide variables and methods that will help to construct other modules.
Math
Just because we are computing stuff, it does not means that we do not need math. Math is the mathematical package, containing low-level math implementations. From random numbers to distributions generation, you can find your needs in this module.
Models
Each machine learning OPF-based technique is defined in this package. From Supervised OPF to Unsupervised OPF, you can use whatever suits your needs.
Stream
Every pipeline has its first step, right? The stream package serves as primary methods to load data, parse it into feasible arrays, and split them into the desired sets (training, evaluation, testing).
Subgraphs
As each type of OPF classifier might need a particular subgraph, we opted to define a unique package to hold them all. The subgraph's package is in charge of implementing some of its varities, such as the k-nearest neighbour (KNN) subgraph.
Utils
This is a utility package. Common things shared across the application should be implemented here. It is better to implement once and use it as you wish than re-implementing the same thing over and over again.
Installation
We believe that everything has to be easy. Not tricky or daunting, OPFython will be the one-to-go package that you will need, from the very first installation to the daily-tasks implementing needs. If you may just run the following under your most preferred Python environment (raw, conda, virtualenv, whatever):
pip install opfython
Alternatively, if you prefer to install the bleeding-edge version, please clone this repository and use:
pip install -e .
Environment configuration
Note that sometimes, there is a need for additional implementation. If needed, from here you will be the one to know all of its details.
Ubuntu
No specific additional commands needed.
Windows
No specific additional commands needed.
MacOS
No specific additional commands needed.
Support
We know that we do our best, but it is inevitable to acknowledge that we make mistakes. If you ever need to report a bug, report a problem, talk to us, please do so! We will be available at our bests at this repository or [email protected].
![[BUG]](https://avatars.githubusercontent.com/u/76575769?v=4)
![[REG] The labels should start from 1, which is not that obvious](https://avatars.githubusercontent.com/u/22035498?v=4)
![[NEW]](https://avatars.githubusercontent.com/u/28956902?v=4)
85 Sep 15, 2022
411 Dec 26, 2022
13.8k Jan 05, 2023
9.3k Jan 07, 2023
43 Dec 04, 2022
50 Dec 29, 2022
32 Jul 19, 2022
141 Jan 07, 2023
66 Nov 16, 2022
47 Dec 06, 2022
160 Jan 04, 2023
129 Jan 04, 2023
245 Dec 20, 2022
90 Sep 10, 2021
48 Dec 20, 2022
101 Sep 07, 2022
348 Dec 22, 2022
6 Sep 10, 2022
14 Nov 24, 2022