from sklearn.model_selection import KFold
from sklearn import tree

def computeCVAccuracy(X,y,folds):
    accuracy=[]
    foldAcc=[]
    depth=[None, 2, 3, 4, 5, 6, 7]
    for i in depth: 
        kf = KFold(n_splits=folds)
        kf.get_n_splits(X)
        print("depth: ", i)
        for train_index, test_index in kf.split(X):
            X_train3, X_test3, Y_train3, Y_test3 = train_test_split(X, Y, train_size = 80, test_size = 20)
            
            clf3 = DecisionTreeClassifier(criterion="entropy", max_depth = i) #, max_depth = 3
            clf3 = clf3.fit(X_train3, Y_train3)

            

            score = clf3.score(X_test3, Y_test3)
            accuracy.append(score)    
        #Building the decision tree classifiers
        dot_data3 = export_graphviz(clf3, out_file=None, 
                                    feature_names = X.columns,  
                                    filled=True, rounded=True,  
                                    special_characters=True)  
        graph = graphviz.Source(dot_data3,  format="png")  
        fig, ax = plt.subplots(figsize=(10, 10))
        tree.plot_tree(clf3, fontsize=10)
        plt.show() 
        #Print Accuracy
        print("Accuracy: ", np.mean(accuracy))

        print("\n")
        foldAcc.append(np.mean(accuracy))  
    return(foldAcc)

computeCVAccuracy(X, Y, 10)