fuzzyops 1.0.16

Fuzzy math library

FuzzyOps

Library of algorithms for fuzzy forecasting and decision support

The library is intended for use:

• in scientific laboratories engaged in research in the field of multi-criteria analysis, optimal planning and management;
• in companies engaged in the development of decision support systems. In fact, the library should be used in the creation of both full-featured software products and experimental mock-ups of software systems designed to work with fuzzy factors.

The library can also be used by directly calling functions in C++ programs, following the instructions:

• https://github.com/Kotya2001/FuzzyOps/blob/main/cpp/README.md

It is also possible to implement a RESTful API in your software (the web service is deployed on your own resources), following these instructions:

• https://github.com/Kotya2001/FuzzyOps-App - the source code of the web service for deployment;
• https://github.com/Kotya2001/FuzzyOps-App/tree/main/posters - examples for accessing a web service via API (implemented in Python);
• https://github.com/Kotya2001/FuzzyOps-App/wiki/Инструкция-по-использованию-алгоритмов-в-веб%E2%80%90серсиве-(по-API) - Instructions for using the library's algorithms through the RESTful API.

How to install the library

To install the library as a pip package, use the command: pip install git+https://{login}:{token}@github.com/Kotya2001/FuzzyOps.git by substituting the appropriate values:

• login: your login on GitHub
• token: how to create a token - here

Or

• pip install fuzzyops

Before installation

Create a virtual environment with Python >= 3.10

Full path to the Python 3.10 executable file -m venv env

Activating the environment

• Macos: source env/bin/activate
• Windows: .\env\Scripts\activate
• Linux: source env/bin/activate

Installing the Cuda Toolkit 11.5

• https://developer.nvidia.com/cuda-11-5-0-download-archive

Install PyTorch depending on your operating system

• Windows: pip3 install torch --index-url https://download.pytorch.org/whl/cu117
• Macos: pip3 install torch
• Linux: pip3 install torch

Minimum technical requirements

• RAM capacity of at least 2 GB;
• For CUDA calculations, an Nvidia GeForce RTX 3060 or higher graphics output device
• Installed Python version 3.10 or higher

Instructions for using the library and documentation for the library's source code:

• Instructions for working with the library - https://github.com/Kotya2001/FuzzyOps/wiki/Instructions-for-using-the-FuzzyOps-library;
• Documentation for the library source code - https://fuzzyops.readthedocs.io/en/latest/

Running tests

After installation, the tests are run according to the instructions.:

• Instructions for running tests - https://github.com/Kotya2001/FuzzyOps/wiki/Instructions-for-running-FuzzyOps-library-tests

Instructions for using the library in C++ programs

• Instructions for using the library in C++ programs - https://github.com/Kotya2001/FuzzyOps/blob/main/cpp/README.md

A web service for accessing the library's algorithms using the RESTful API:

• The source code of the web service and instructions for using it - https://github.com/Kotya2001/FuzzyOps-App

Description of folders with library repository files

• cpp - Instructions for using the library in C++ programs and examples of using the library in Python and C++;
• example:
  • common - Examples of using the library code;
  • The remaining files are practical examples of using the library code;
• src - Library source codes:
  • docs - Files, format .html with documentation for the source code (compiled using the library sphinx);
  • fuzzyops - Library source codes:
    • fan - Source codes of fuzzy analytical networks;
    • fuzzy_logic Source codes of fuzzy logic algorithms;
    • fuzzy_msa - Source codes of classical multicriteria analysis algorithms with fuzzy variables;
    • fuzzy_nn - Source codes of algorithms for fuzzy neural networks (ANFIS Network);
    • fuzzy_numbers - Source codes for implementing fuzzy numbers (fuzzification, defuzzification, fuzzy arithmetic);
    • fuzzygraphs - Source codes for the implementation of fuzzy graphs;
    • fuzzygraphs_algs - Source codes of algorithms on fuzzy graphs (Fuzzy dominance relations, fuzzy factor models, fuzzy transport graphs);
    • fuzzy_pred - Source codes of fuzzy prediction algorithms;
    • sequencing_assignment - The source codes of algorithms on fuzzy graphs of the sequence of work in assignment tasks;
    • tests - Algorithm Test Codes.
• readthedocs - A file for automatic assembly and placement of documentation on https://about.readthedocs.com;
• doc_reqs.txt - A library dependency file for building documentation https://about.readthedocs.com;
• requirements - The dependency file for installing the library;
• setup.cfg - Configuration file for building the library distribution;
• setup.py - A file for building a library distribution using setuptools;
• LICENSE - Library license file;