https://github.com/olivervea/fluentfuzzy

FluentFuzzy is a package for doing fuzzy logic in .NET with a natural fluent syntax.
https://github.com/olivervea/fluentfuzzy

fuzzy-logic nuget unity winforms

Last synced: 4 days ago
JSON representation

FluentFuzzy is a package for doing fuzzy logic in .NET with a natural fluent syntax.

• Host: GitHub
• URL: https://github.com/olivervea/fluentfuzzy
• Owner: OliverVea
• License: mit
• Created: 2023-08-17T18:15:18.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-12-09T06:34:37.000Z (2 months ago)
• Last Synced: 2024-12-12T20:16:45.860Z (about 2 months ago)
• Topics: fuzzy-logic, nuget, unity, winforms
• Language: C#
• Homepage:
• Size: 357 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# FluentFuzzy

FluentFuzzy is a package for doing fuzzy logic in .NET with a natural fluent syntax.

> FluentFuzzy is very much WIP and should be expected to be updated with breaking changes.

## Contents

`FluentFuzzy` contains the core fuzzy logic functionality.

`FluentFuzzy.Visualization` has visualization for some `FluentFuzzy` classes.

## Using

After importing the Nuget package, using FluentFuzzy is a 2-step process.

### Setting up input and output

`FuzzyInput`s and `FuzzyOutput`s can be set up in the following manner:

```cs

// test/FluentFuzzy.Test/Example.cs#L10-L12

const int healthValue = 35;

var health = new FuzzyInput(() => healthValue);

var flee = new FuzzyOutput();

```

`FuzzyInput` requires a `Func` to get the current crisp value of the input.

### Setting up member functions

The next step is adding member functions to each input:

```cs

// test/FluentFuzzy.Test/Example.cs#L14-L20

const int low = 0;

const int medium = 1;

const int high = 2;

health.Set(low, new Trapezoid(0, 0, 25, 50));

health.Set(medium, new Triangle(25, 50, 75));

health.Set(high, new Trapezoid(50, 75, 100, 100));

```

Currently, the following membership functions have been implemented:

* `Triangle` defines a triangle.

* `Trapezoid` defines a trapezoid.

* `Line` defines a line.

* `Value` defines a single value.

* `AtLeast` defines a threshold value.

Custom membership functions can be created by implementing the `IMembershipFunction` interface.

After the input functions, membership functions can be added to the `FuzzyOutput`

```cs

// test/FluentFuzzy.Test/Example.cs#L22-L24

flee.Set(low, new Triangle(-0.5, 0, 0.5));

flee.Set(medium, new Triangle(0, 0.5, 1));

flee.Set(high, new Triangle(0.5, 1, 1.5));

```

The only currently supported defuzzification method is a version of the centroid method. As some of the membership functions, namely `Line`, `Value` and `AtLeast` do not have useful centroids, these cannot be used as output member functions.

Custom output membership functions can be created by implementing the `IHasCentroid` interface.

### Setting up fuzzy rules

After setting up the fuzzy inputs and outputs, rules can be created to connect them.

Rules are created with a readable fluent syntax:

```cs

// test/FluentFuzzy.Test/Example.cs#L26-L28

FuzzyRule.If(health.Is(high)).Then(flee.Is(low));

FuzzyRule.If(health.Is(medium)).Then(flee.Is(medium));

FuzzyRule.If(health.Is(low)).Then(flee.Is(high));

```

### Evaluating an output

After creating the fuzzy ruleset, the output can be evaluated by calling the `Evaluate` method on the output:

```cs

// test/FluentFuzzy.Test/Example.cs#L30-L30

Console.WriteLine($"flee(health: {health.Value}) = {flee.Evaluate()}"); // flee(health: 35) = 0,8

```