FuzzyInfer¶

Unix-compliant fuzzy logic inference with a beautiful fluent API.

Why FuzzyInfer?¶

Real-world reasoning is rarely black and white. FuzzyInfer brings the power of fuzzy logic to Python with:

🔧 Unix Philosophy: Composable tools that do one thing well
🐍 Pythonic API: Beautiful, fluent interface that feels natural
🚀 Stream-First: Process infinite streams of facts and rules
📊 Degrees of Belief: Express uncertainty with values between 0.0 and 1.0
🔗 Pipeline Ready: Chain with other Unix tools via JSONL

Quick Example¶

Fluent APIUnix PipelinePattern Matching

from fuzzy_infer import fuzzy

# Beautiful, chainable inference
results = (
    fuzzy()
    .fact("is-bird", ["robin"], 0.9)
    .fact("has-wings", ["robin"], 1.0)
    .rule(
        when=("is-bird", "?x"),
        then=("can-fly", "?x", 0.85)
    )
    .infer()
    .query("can-fly", ["robin"])
)
# Result: can-fly(robin) with degree 0.765

# Stream facts through inference pipeline
echo '{"type":"fact","pred":"is-bird","args":["robin"],"degree":0.9}' |
fuzzy-infer run rules.jsonl |
fuzzy-infer query "can-fly"

# Powerful pattern variables
fuzzy().rule(
    when=[
        ("parent", "?x", "?y"),    # ?x is parent of ?y
        ("parent", "?y", "?z")     # ?y is parent of ?z
    ],
    then=("grandparent", "?x", "?z")  # ?x is grandparent of ?z
)

Key Features¶

🎯 Forward-Chaining Inference¶

Automatically derive new facts from rules until no new knowledge can be inferred.

🔄 Stream Processing¶

Process infinite streams of facts and rules - perfect for real-time systems.

🎨 Fluent API¶

Chain operations naturally with Python's most beautiful inference API.

🐧 Unix Composability¶

Works seamlessly with grep, jq, awk, and other Unix tools via JSONL.

📐 Pattern Variables¶

Express complex relationships with pattern matching and variable binding.

💫 Fuzzy Operations¶

Built-in fuzzy logic operators (AND, OR, NOT) with configurable T-norms.

Installation¶

pip install fuzzy-infer

Or for development:

git clone https://github.com/queelius/fuzzy-infer.git
cd fuzzy-infer
pip install -e ".[dev]"

Next Steps¶

Quick Start

Get up and running in 5 minutes

Quick Start Guide
Learn Concepts

Understand fuzzy logic and inference

Core Concepts
Fluent API

Master the Pythonic interface

API Guide
Unix CLI

Build inference pipelines

CLI Usage

Philosophy¶

FuzzyInfer embraces the Unix philosophy:

Do One Thing Well: Fuzzy inference, nothing more
Text Streams: JSONL for universal compatibility
Composable: Chain with any tool that speaks JSON
Simple: Clear, minimal API that's easy to learn

License¶

MIT License - see LICENSE for details.