Variables & rules¶

Linguistic variables¶

import fuzzytool as fz

# Auto-generate evenly-spaced terms ("triangular" default, or "gauss"):
temp = fz.Variable("temp", (0, 40), terms=["cold", "warm", "hot"], kind="gauss")

# Or assign terms explicitly:
temp["freezing"] = fz.trap(-10, -10, 0, 5)

Rules with operators¶

Indexing a variable yields a proposition (temp["hot"]). Compose propositions:

Operator	Meaning	Default
`&`	AND	t-norm `min`
`\\|`	OR	s-norm `max`
`~`	NOT	complement `1 - μ`

fan = fz.Variable("fan", (0, 100), terms=["off", "low", "high"])

sys = fz.Mamdani()
sys.rule(temp["hot"] & ~temp["cold"], fan["high"])
sys.rule(temp["warm"], fan["low"])

Each rule may carry a weight in [0, 1]:

sys.rule(temp["cold"], fan["off"], weight=0.5)

Choosing connectives¶

Pass them to the engine by name (or supply your own callable):

sys = fz.Mamdani(tnorm="prod", snorm="probor")

See fuzzytool.norms for the available t-norms and s-norms.