from sklearn.tree import DecisionTreeClassifierimport pandas as pdfrom sklearn.metrics import accuracy_score## training data (20%)data = pd.read_csv("train.csv", usecols=[1,2,9])X_train = pd.read_csv("train.csv", usecols=[2,9])y_train = pd.read_csv("train.csv", usecols=[1])dt = DecisionTreeClassifier(max_depth=6)dt.fit(X_train, y_train)y_predict = dt.predict(X_test)accuracy = dt(y_test, y_predict)我得到“TypeError：'DecisionTreeClassifier'對象不可調(diào)用”，即使我（主要遵循datacamp教程）。
1 回答
月關(guān)寶盒
TA貢獻(xiàn)1772條經(jīng)驗(yàn) 獲得超5個(gè)贊
dt(...)試圖“調(diào)用”，你不能這樣做，因?yàn)椴皇且粋€(gè)函數(shù)。您需要一個(gè)從真實(shí)標(biāo)簽和預(yù)測標(biāo)簽計(jì)算準(zhǔn)確性的函數(shù)。dtdt
試試這樣的東西
def calculate_accuracy(y_true, y_predicted):
num_correct = sum(map(lambda t, p: t==p, y_true, y_predicted))
return num_correct / len(y_true)
accuracy = calculate_accuracy(y_test, y_predict)
編輯：
我的更初學(xué)者友好的版本看起來像calculate_accuracy()
def calculate_accuracy(y_true, y_predicted):
num_correct = 0
for i in range(len(y_true)):
if (y_true[i] == y_predicted[i]):
num_correct = num_correct + 1
return num_correct / len(y_true)
添加回答
舉報(bào)
0/150
提交
取消