FuzzyPy

The library for working with fuzzy logic in python.

This library was developed with simplicity and independability in mind. The main idea was to use functional approach as the basis of the fuzzy logic is membership functions combinations and Python has powerful functional conception.

What is this repository for?

This library is developed to be used for modelling the fuzzy sets and fuzzy controllers. The library cab be used both in production and for educational purposes.

Version 1.0

The example

from fuzzypy.defuzzification import defuzzify
from fuzzypy.memberships import TriFunc, TrapecFunc
from fuzzypy.variables import FuzzyVariable, FuzzyRule, build_resulting_fuzzy_term

if __name__ == "__main__":
    # Create a fuzzy variable
    fuzzy_temp = FuzzyVariable()  # Temperature

    # Define the membership functions
    hot = TriFunc(20, 25, 50)
    norm = TriFunc(15, 20, 25)
    cold = TrapecFunc(0, 5, 10, 20)

    # Determine the fuzzy terms
    temp_is_hot = fuzzy_temp.is_(hot)  # The temperature is hot
    temp_is_norm = fuzzy_temp.is_(norm)  # The temperature is normal
    temp_is_cold = fuzzy_temp.is_(cold)  # The temperature is cold

    # Create an output fuzzy variable
    fuzzy_blow = FuzzyVariable()  # The speed of the fan

    # and its membership functions
    slow = TriFunc(0, 0, 750)
    fast = TriFunc(250, 1000, 1000)

    # Determine the rules
    rules = [
        FuzzyRule(temp_is_cold | temp_is_norm, fuzzy_blow, slow),  # If the temperature is cold or normal then fan speed is slow
        FuzzyRule(temp_is_hot, fuzzy_blow, fast),  # If the temperature is hot then fan speed is fast
    ]

    fuzzy_temp.value = 30  # Let the temperature be 30 degrees

    # Let's find the limits of the variables
    print("Temp lower limit is {}".format(fuzzy_temp.low_limit))
    print("Temp upper limit is {}".format(fuzzy_temp.upp_limit))
    print("Blow lower limit is {}".format(fuzzy_blow.low_limit))
    print("Blow upper limit is {}".format(fuzzy_blow.upp_limit))

    print("The temperature is {}".format(fuzzy_temp.value))
    fuzzy_fan_speed = build_resulting_fuzzy_term(rules, fuzzy_blow)
    fan_speed = defuzzify(fuzzy_fan_speed)  # the default method is center-of-gravity

    print("Defuzzyfied values are {}".format(fan_speed))  # it could be more then one value
    print("Or defuzzyfied fan speed is {}".format(fuzzy_blow.value))  # and we can check the value of the fan speed

    # directly
    fuzzy_temp.value = 14  # and now the temp is 13 degrees
    fuzzy_fan_speed = build_resulting_fuzzy_term(rules, fuzzy_blow)
    defuzzify(fuzzy_fan_speed)  # no explicit defuzzified value, but we have it in the variable
    print("Now the temperature is {}, and the fan speed is {}".format(fuzzy_temp.value, fuzzy_blow.value))  # and print it

How do I get set up?

• To setup the packet use pip: pip install fuzzypy
• No configuration is needed.
• Dependencies: see requirements.txt. At the moment no external dependencies are required except for testing purposes.
• How to run the tests: in the project folder run command

python setup.py test

Or you can use the following commands:

python -m unittest tests/Implication_test.py tests/membership_functions_test.py tests/variables_test.py tests/defuzzify_test.py
coverage run -m unittest tests/Implication_test.py tests/membership_functions_test.py tests/variables_test.py tests/defuzzify_test.py
coverage report -m

Contribution

The contribution would be highly appreciable.

Now it is good to implement another defuzzification methods (see defuzzification methods)