fuzzy4
A Python library for 4-valued fuzzy logic based on Belnap logic with product t-norm operations.
Overview
Traditional boolean logic has two values: True and False. Real-world knowledge often involves uncertainty (we don't know) and contradiction (conflicting evidence). Belnap's 4-valued logic addresses this by representing propositions as vectors (T, F):
| State | Vector | Meaning |
|---|---|---|
| TRUE | (1, 0) | Confirmed, not refuted |
| FALSE | (0, 1) | Refuted, not confirmed |
| UNKNOWN | (0, 0) | Neither confirmed nor refuted |
| CONFLICT | (1, 1) | Both confirmed and refuted |
This library extends Belnap logic with fuzzy values, where T and F are continuous values in [0, 1].
Installation
pip install fuzzy4
Quick Start
from fuzzy4 import FuzzyBool, TRUE, FALSE, UNKNOWN, CONFLICT
# Create fuzzy values
x = FuzzyBool(0.8, 0.2) # 80% confirmed, 20% refuted
y = FuzzyBool(0.6, 0.3) # 60% confirmed, 30% refuted
# Logical operations
result = x & y # conjunction (AND)
result = x | y # disjunction (OR)
result = ~x # negation (NOT)
result = x >> y # implication (x -> y)
result = x.iff(y) # bi-implication (x <-> y)
# Accumulate evidence
result = x + y # combines evidence from multiple sources
result += FuzzyBool(0.5, 0.1) # in-place accumulation
result = FuzzyBool.accumulate(x, y, z) # accumulate many values
# Check dominant state
print(x.dominant_state()) # "true", "false", "unknown", or "conflict"
print(x.is_true()) # True if predominantly true
Operations
Negation: ~x
Swaps confirmation and refutation:
| x | ~x |
|---|---|
| TRUE | FALSE |
| UNKNOWN | UNKNOWN |
| CONFLICT | CONFLICT |
| FALSE | TRUE |
~FuzzyBool(0.8, 0.3) # -> FuzzyBool(0.3, 0.8)
Conjunction: x & y
Both must confirm, any refutation propagates:
| & | TRUE | UNKNOWN | CONFLICT | FALSE |
|---|---|---|---|---|
| TRUE | TRUE | UNKNOWN | CONFLICT | FALSE |
| UNKNOWN | UNKNOWN | UNKNOWN | FALSE | FALSE |
| CONFLICT | CONFLICT | FALSE | CONFLICT | FALSE |
| FALSE | FALSE | FALSE | FALSE | FALSE |
Uses product t-norm for T and probabilistic s-norm for F:
T = Tx * TyF = Fx + Fy - Fx * Fy
Disjunction: x | y
Any confirmation suffices, both must refute:
| | | TRUE | UNKNOWN | CONFLICT | FALSE |
|---|---|---|---|---|
| TRUE | TRUE | TRUE | TRUE | TRUE |
| UNKNOWN | TRUE | UNKNOWN | TRUE | UNKNOWN |
| CONFLICT | TRUE | TRUE | CONFLICT | CONFLICT |
| FALSE | TRUE | UNKNOWN | CONFLICT | FALSE |
Uses probabilistic s-norm for T and product t-norm for F:
T = Tx + Ty - Tx * TyF = Fx * Fy
Implication: x >> y
Material implication: x -> y = ~x | y
| >> | TRUE | UNKNOWN | CONFLICT | FALSE |
|---|---|---|---|---|
| TRUE | TRUE | UNKNOWN | CONFLICT | FALSE |
| UNKNOWN | TRUE | UNKNOWN | TRUE | UNKNOWN |
| CONFLICT | TRUE | TRUE | CONFLICT | CONFLICT |
| FALSE | TRUE | TRUE | TRUE | TRUE |
Bi-implication: x.iff(y)
Equivalence: x <-> y = (x -> y) & (y -> x)
| iff | TRUE | UNKNOWN | CONFLICT | FALSE |
|---|---|---|---|---|
| TRUE | TRUE | UNKNOWN | CONFLICT | FALSE |
| UNKNOWN | UNKNOWN | UNKNOWN | TRUE | UNKNOWN |
| CONFLICT | CONFLICT | TRUE | CONFLICT | CONFLICT |
| FALSE | FALSE | UNKNOWN | CONFLICT | TRUE |
Evidence Accumulation: x + y
Combines evidence using probabilistic sum — useful for aggregating information from multiple sources:
| + | TRUE | UNKNOWN | CONFLICT | FALSE |
|---|---|---|---|---|
| TRUE | TRUE | TRUE | CONFLICT | CONFLICT |
| UNKNOWN | TRUE | UNKNOWN | CONFLICT | FALSE |
| CONFLICT | CONFLICT | CONFLICT | CONFLICT | CONFLICT |
| FALSE | CONFLICT | FALSE | CONFLICT | FALSE |
weak = FuzzyBool(0.3, 0.0) # weak confirmation
result = weak + weak + weak # -> FuzzyBool(0.657, 0.0)
# In-place accumulation
total = UNKNOWN
for source in sources:
total += source # accumulates evidence iteratively
# Accumulate a list of values
FuzzyBool.accumulate([source1, source2, source3])
Properties:
- More confirmations → T approaches 1
- More refutations → F approaches 1
- UNKNOWN is the neutral element
Scalar Operations
x = FuzzyBool(0.8, 0.6)
x * 0.5 # -> FuzzyBool(0.4, 0.3)
x / 2.0 # -> FuzzyBool(0.4, 0.3)
Component Extraction
Each FuzzyBool has four derived properties representing the degree of each logical state:
x = FuzzyBool(0.7, 0.4)
x.truth # degree of pure truth (confirmed AND not refuted)
x.falsity # degree of pure falsity (refuted AND not confirmed)
x.unknown # degree of uncertainty (neither confirmed nor refuted)
x.conflict # degree of contradiction (both confirmed and refuted)
x.as_dict() # {"truth": ..., "falsity": ..., "unknown": ..., "conflict": ...}
Use Cases
- Knowledge bases: Handle incomplete or contradictory information
- Sensor fusion: Combine readings from multiple sensors with varying reliability
- Expert systems: Model uncertainty in rule-based reasoning
- Database queries: Represent null/unknown values with more nuance than SQL's 3-valued logic
Mathematical Background
This implementation uses:
- Product t-norm for conjunction:
a ⊗ b = a × b - Probabilistic s-norm for disjunction:
a ⊕ b = a + b - a × b - Standard negation:
¬a = 1 - a
De Morgan's laws hold: ~(x & y) = ~x | ~y and ~(x | y) = ~x & ~y
Development
# Clone the repository
git clone https://github.com/mihail-gribov/fuzzy4.git
cd fuzzy4
# Install with dev dependencies
pip install -e ".[dev]"
# Run tests
pytest
# Run tests with coverage
pytest --cov=fuzzy4
License
MIT License — see LICENSE for details.
References
- Belnap, N. D. (1977). "A useful four-valued logic"
- Fitting, M. (1994). "Kleene's Three Valued Logics and Their Children"
Metadata
Release files for fuzzy4 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| fuzzy4-0.1.0.tar.gz | 9.4 kB | Without provenance | Details for fuzzy4-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| fuzzy4-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for fuzzy4-0.1.0-py3-none-any.whl |
Total release size: 17.3 kB
Release files / fuzzy4-0.1.0.tar.gz
| Download URL | fuzzy4-0.1.0.tar.gz |
|---|---|
| Size | 9.4 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
b5be99e255a1802e2d26de2749ef188ffa81755a0bc30791151906659c92b10e
|
|
BLAKE2b-256 checksum How to use checksums |
9c72eac1eded5b2b87766d8fe37ae8d12c5842a1011db647caa29dd53ed180fc
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.11
|
Release files / fuzzy4-0.1.0-py3-none-any.whl
| Download URL | fuzzy4-0.1.0-py3-none-any.whl |
|---|---|
| Size | 7.8 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
85c053a9cf147055126770461c2599dcf5d76b34dd6c1a53924f6c103c726476
|
|
BLAKE2b-256 checksum How to use checksums |
006be000266c1f83952c7c8dc03fed5373545176be031b53508c0403b42bce07
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.11
|
Security for fuzzy4 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.