Fix: output variable can set default values when the output is a single scalar (#92)

Author: jcrada

fuzzylite/variable.py

@@ -532,7 +532,8 @@ def defuzzify(self) -> None:
 
         # Applying default values
         if not np.isnan(self.default_value):
-            value[np.isnan(value)] = self.default_value  # type: ignore
+            value = np.atleast_1d(value)
+            value[np.isnan(value)] = self.default_value
 
         # Committing the value
         self.value = value

tests/test_issues.py

@@ -0,0 +1,86 @@
+import fuzzylite as fl
+import unittest
+import numpy as np
+
+
+class TestIssue91(unittest.TestCase):
+    """
+    Test case for https://github.com/fuzzylite/pyfuzzylite/issues/91
+    """
+
+    def engine(self) -> fl.Engine:
+        fll = """
+Engine: tipper
+description: (service and food) -> (tip)
+InputVariable: service
+  description: quality of service
+  enabled: true
+  range: 0.000 10.000
+  lock-range: true
+  term: poor Discrete 0.000 1.000 2.500 1.000 5.000 0.000 10.000 0.000
+  term: good Discrete 0.000 0.000 2.500 0.000 5.000 1.000 7.500 0.000 10.000 0.000
+  term: excellent Discrete 0.000 0.000 5.000 0.000 7.500 1.000 10.000 1.000
+InputVariable: food
+  description: quality of food
+  enabled: true
+  range: 0.000 10.000
+  lock-range: true
+  term: rancid Discrete 0.000 1.000 2.500 1.000 7.500 0.000 10.000 0.000
+  term: delicious Discrete 0.000 0.000 2.500 0.000 7.500 1.000 10.000 1.000
+OutputVariable: mTip
+  description: tip based on Mamdani inference
+  enabled: true
+  range: 0.000 30.000
+  lock-range: false
+  aggregation: Maximum
+  defuzzifier: Centroid 100
+  default: 0.0
+  lock-previous: false
+  term: cheap Discrete 0.000 0.000 5.000 1.000 10.000 0.000 30.000 0.000
+  term: average Discrete 0.000 0.000 10.000 0.000 15.000 1.000 20.000 0.000 30.000 0.000
+  term: generous Discrete 0.000 0.000 20.000 0.000 25.000 1.000 30.000 0.000
+OutputVariable: tsTip
+  description: tip based on Takagi-Sugeno inference
+  enabled: true
+  range: 0.000 30.000
+  lock-range: false
+  aggregation: none
+  defuzzifier: WeightedAverage TakagiSugeno
+  default: 0.0
+  lock-previous: false
+  term: cheap Constant 5.000
+  term: average Constant 15.000
+  term: generous Constant 25.000
+RuleBlock: mamdani
+  description: Mamdani inference
+  enabled: true
+  conjunction: AlgebraicProduct
+  disjunction: AlgebraicSum
+  implication: Minimum
+  activation: General
+  rule: if service is poor or food is rancid then mTip is cheap
+  rule: if service is good then mTip is average
+  rule: if service is excellent or food is delicious then mTip is generous with 0.5
+  rule: if service is excellent and food is delicious then mTip is generous with 1.0
+RuleBlock: takagiSugeno
+  description: Takagi-Sugeno inference
+  enabled: true
+  conjunction: AlgebraicProduct
+  disjunction: AlgebraicSum
+  implication: none
+  activation: General
+  rule: if service is poor or food is rancid then tsTip is cheap
+  rule: if service is good then tsTip is average
+  rule: if service is excellent or food is delicious then tsTip is generous with 0.5
+  rule: if service is excellent and food is delicious then tsTip is generous with 1.0
+        """
+        return fl.FllImporter().from_string(fll)
+
+    def test_issue_91(self) -> None:
+        # "'numpy.float64' object does not support item assignment"
+        engine = self.engine()
+        engine.variable("food").value = 10
+        engine.variable("service").value = 10
+        engine.process()
+        np.testing.assert_almost_equal(engine.variable("mTip").value, 25.00104979)
+        np.testing.assert_almost_equal(engine.variable("tsTip").value, 25)