Overview

pyFUZZYgenerator is a framework to automatically generate the code to evaluate a fuzzy system in ANSI C. It uses configuration files to define the characteristics of the fuzzy system.

Usage

To generate a fuzzy library, a configuration file is required. Configuration files are in TOML format.

In the examples directory there is an example configuration file that includes all the available fuzzy types.

In the following, the default example for the F-IND model examples/sampleFIND.toml will be used.

To create the set of C source files using the sample file, run the main.py script as follows:

python3 main.py examples/sampleFIND.toml -d sample

The generated C source files composing the fuzzy evaluator are placed in the sample directory.

More details on the options and arguments can be obtained with the command:

python3 main.py --help

The generated source files include a Makefile, so that the executable can be generated as follows:

cd sample
make

These commands generated the sample executable that can be ran with

./sample

If the model is run without the necessary input parameters, a warning is printed on standard error, and the model uses default values equal to 0 for all the input variables.

The input values for the variables of a model can be specified on the command line as follows:

./sample --Test=0.5,1,0.3

This means that the model Test requires 3 input values, which are 3 comma-separated numerical values.

The generated fuzzy evaluator supports basic help:

./sample --help
Usage: test [OPTION...]

      --Test=VALUES          Input of model Test: A B C 
  -?, --help                 Give this help list
      --usage                Give a short usage message

Dependencies

The Python (python3) library requires the following packages:

• pytoml
• jinja2
• numpy

To simplify the requirements installation, we provide an Anaconda environment, that can be imported with:

conda env create -f environment.yaml

TODOs

• Implement the C evaluator for a general fuzzy system
• Optimize the general fuzzy library
• Make pyFUZZYgenerator a program installable with pip