Sample mean in RandomForest-CodePudding

i did taking max class. How to change predict method in Leaf class from taking the maximum class to taking the sample mean?


        prediction = max(classes, key=classes.get)

CodePudding user response：

I think this should make the current code testable. I've made some assumptions, I hope they're correct:

class Leaf:
    
    def __init__(self, data, labels):
        self.data = data
        self.labels = labels
        self.prediction = self.predict()
        
    def predict(self):
        classes = {} 
        for label in self.labels:
            if label not in classes:
                classes[label] = 0
            classes[label]  = 1
              
        prediction = max(classes, key=classes.get)
        return prediction   
print(Leaf(None, ["a", "b", "b"]).prediction, "is the most common leaf")

Ouput:

b is the most common leaf

So, I guess you want to find the median leaf? I don't know for average leaf.

class Leaf:
    def __init__(self, data, labels):
        self.data = data
        self.labels = labels
        self.prediction = self.predict()
    def predict(self):
        classes = {} 
        for label in self.labels:
            if label not in classes:
                classes[label] = 0
            classes[label]  = 1      
        prediction = sorted(classes, key=classes.get)[len(classes) // 2]
        return prediction
assert Leaf(None, ["a"]).prediction == "a"
assert Leaf(None, ["a", "a"]).prediction == "a"
assert Leaf(None, ["a", "b"]).prediction == "b"
assert Leaf(None, ["a", "a", "a"]).prediction == "a"
assert Leaf(None, ["a", "a", "b"]).prediction == "a"
assert Leaf(None, ["a", "b", "c"]).prediction == "b"
assert Leaf(None, ["a", "a", "a", "a"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "b"]).prediction == "a"
assert Leaf(None, ["a", "a", "b", "b"]).prediction == "b"
assert Leaf(None, ["a", "a", "b", "c"]).prediction == "c"
assert Leaf(None, ["a", "a", "a", "a", "a"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "a", "b"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "b", "b"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "b", "c"]).prediction == "c"
assert Leaf(None, ["a", "a", "b", "b", "b"]).prediction == "b"
assert Leaf(None, ["a", "a", "b", "b", "c"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "a", "a", "a"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "a", "a", "b"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "a", "b", "b"]).prediction == "a"
assert Leaf(None, ["a", "a", "a", "a", "b", "c"]).prediction == "c"
assert Leaf(None, ["a", "a", "a", "b", "b", "b"]).prediction == "b"
assert Leaf(None, ["a", "a", "a", "b", "b", "c"]).prediction == "b"
assert Leaf(None, ["a", "a", "b", "b", "c", "c"]).prediction == "b"