Skip to main content

A package which implements various systems of n valued logic.

Project description

MVL (many valued logic)

MVL (many valued logic) is a flexible, expressive, and extensible python package which makes it easy to work with many valued logic systems: logic systems which use more than 2, or infinite truth values, beyond «True» and «False».

Latest release 0.2.0

Table of contents

Problem statement

By default, python doesn't provide any boolean infrastructure for values other than «True» and «False». This makes it verbose, difficult, or impossible to deal with situations involving many valued logic.

Example

To demonstrate this limitation, we'll use an example in logic. Consider the sentence:

The apple is red.

We have an apple that's kinda redish yellow, or somewhere in the middle. So the answer to this question is a very half-hearted «maybe?».

Now consider the sentence:

The apple is red, and the apple is not red.

In 2 valued logic, this would always be false. But our apple is maybe red, but also maybe not red.

3 valued logic has a solution to problems like this, where we have ranges of truth values between true and false. Say our answer to the first question is maybe then our answer to the second question can be maybe and maybe. So our final answer to the second question is maybe.

In real life, SQL uses 3 valued logic to implement its logic. MVL provides an easy way to implement these sort of checks in python.

What does MVL do?

MVL provides classes and functions which implement 3, n, or ∞ valued logic which can be imported and used in code bases. It provides conversion between floats and logical values to allow different degrees of truthfulness, provides a library of default operators, and allows for the implementation of custom operators which can be used to analyze logical statements of arbitrary values.

Example usage

Going back to our apple example, to do logic on keys that may be pressed, we need more than just «true» and «false» values to deal with this -- we need 3 valued logic.

Writing a logical condition for the second sentence will look like this:

if (priest.bool_(
    priest.or_(is_red(apple), priest.not_(is_red(apple)))
)):
    do something

Let's break this down.

First, imagine that is_red(apple) is some function defined somewhere else that returns one of 3 logic values telling us if apple is red.

We're going to use priest logic, which just means that we use 3 boolean values, and that anything that's not «false» will evaluate to «true» when we convert it into python's 2 valued boolean system. The 3 values we use are («priest.f» for «false»; «priest.u» for «unknown»; and «priest.t» for «true»).

if(...) is a normal python if statement. This needs a normal, 2 valued python boolean in order to work. We'll come back to this.

priest.bool_(...) will convert whatever our 3 valued logic value is into a 2 valued python boolean. Exactly how this is defined depends a on the logic system we choose to use. We've used priest logic, so «priest.u» and «priest.t» will evaluate to python's built-in «True», and «priest.f» will evaluate to python's built-in «False».

There's more to this function going on behind the scenes. You won't need it for this example, but you can read about it in the documentation if you're interested.

The rest of the example is logical statements made up of comparisons between the key state (pressed or not pressed) and the priest logic value we want them to be. This will give us a result as a float between 0 and 1.

The result is passed into priest.bool_ to give a python boolean, which is passed into the if statement to decide whether or not to output a capital letter.

Where to get it

The source code is hosted on github at https://github.com/andrewjunyoung/mvl.

PyPI (pip)

To install mvl through pip, open a command line interface and run

$ pip install mvl

Features

Summary

  • Logical systems which can use 3, n, or infinite logic values.
  • A rich library of logical operators for 3, n, and ∞ valued logic.
  • Conversion between floats and logical values.

Supported logic systems

The following 3 valued logic systems are supported by MVL:

  • Bochvar
  • Kleene
  • Priest

The following n valued logic systems are supported by MVL:

  • Łukasiewicz
  • Gödel (under the name «goedel»)
  • Product logic
  • Post logic

The following systems are planned for future support:

  • Belnap's 4 valued logic

Usage examples

Using MVL is designed to integrate with existing python infrastructure as much as possible. Example usages of kleene and lukasiewicz logic are given below.

Importing and using 3 valued systems

>>> from mvl import kleene
>>> kleene.t
LukasiewiczLogicValue.True
>>> kleene.and_(k.t, k.u)
0.5
>>> kleene.or_(k.u, k.u)
0.5
>>> kleene.implies(k.u, k.u)
0.5
>>> kleene.implies(k.f, k.u)
1.0

Creating n valued logical systems

>>> from mvl.lukasiewicz import *
>>> ls = LogicSystem(5, LukasiewiczLogicValue)
>>> ls.values
[LukasiewiczLogicValue(0.0), LukasiewiczLogicValue(0.25), LukasiewiczLogicValue(0.5), LukasiewiczLogicValue(0.75), LukasiewiczLogicValue(1.0)]
>>> t = ls.values[4]
>>> bool(t)
True
>>> bool(ls.values[3])
False

Creating arbitrary logical values

>>> from mvl.lukasiewicz import *
>>> x = LukasiewiczLogicValue(0.123)
>>> x
LukasiewiczLogicValue(0.123)
>>> bool(x)
False
>>> s_and(x, 1)
0.123

License

Unlicense.

Documentation

The sphinx documentation can be generated and opened as follows:

  1. Open a command line.
  2. Navigate to the «docs» directory.
  3. Run the command make html

It can be read by opening the file <project_path>/docs/build/html/index.html.

Links

Resource
How do I know what logic system to use? Todo
Documentation Todo

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for mvl, version 0.2.0
Filename, size File type Python version Upload date Hashes
Filename, size mvl-0.2.0-py3-none-any.whl (14.9 kB) File type Wheel Python version py3 Upload date Hashes View hashes
Filename, size mvl-0.2.0.tar.gz (13.1 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page