import numpy as np
import pandas as pd
from sklearn import tree
import graphviz
# x，y是sklearn中需要擬合的數(shù)據(jù)
x = np.array(exam_train)
y = np.array(classes_train)
clf = tree.DecisionTreeClassifier(criterion='entropy', class_weight='balanced', max_depth=25)
clf = clf.fit(x, y)
dot_data = tree.export_graphviz(clf, out_file=None, feature_names=None, filled=True, rounded=True)  # 重要參數(shù)可定制
graph = graphviz.Source(dot_data)
graph.render(view=True, format="pdf", filename="decisiontree_pdf")

age look income orderly target
older ugly low yes no
young ugly high no no
young handsome low no no
young handsome high yes yes
young handsome medium yes yes
young handsome medium no no

# -*- coding:utf-8*-
# 將字典 轉(zhuǎn)化為 sklearn 用的數(shù)據(jù)形式 數(shù)據(jù)型 矩陣
from sklearn.feature_extraction import DictVectorizer
import csv
from sklearn import preprocessing
from sklearn import tree

allElectronicsData = open('c:/pic/data/tree.csv','rb')
reader = csv.reader(allElectronicsData)
header = reader.next()
# print header
## 數(shù)據(jù)預(yù)處理
featureList = []
labelList = []
for row in reader:
    # print row[-1]
    labelList.append(row[-1])
    # 下面這幾步的目的是為了讓特征值轉(zhuǎn)化成一種字典的形式，就可以調(diào)用sk-learn里面的DictVectorizer，直接將特征的類別值轉(zhuǎn)化成0,1值
    rowDict = {}
    for i in range(1, len(row) - 1):
        rowDict[header[i]] = row[i]
    featureList.append(rowDict)

for each in featureList:
    print each

# Vectorize features
vec = DictVectorizer()
dummyX = vec.fit_transform(featureList).toarray()
print("dummyX:"+str(dummyX))
print(vec.get_feature_names())

# label的轉(zhuǎn)化，直接用preprocessing的LabelBinarizer方法
lb = preprocessing.LabelBinarizer()
dummyY = lb.fit_transform(labelList)
print("dummyY:"+str(dummyY))
print("labelList:"+str(labelList))

#criterion是選擇決策樹節(jié)點的 標準 ，這里是按照“熵”為標準，即ID3算法；默認標準是gini index，即CART算法。
clf = tree.DecisionTreeClassifier()
clf = clf.fit(dummyX,dummyY)
print("clf:"+str(clf))
# 導(dǎo)入相關(guān)函數(shù)，可視化決策樹
# 導(dǎo)出的結(jié)果是一個dot文件(在系統(tǒng)默認路勁)，需要安裝Graphviz才能將它住哪華為PDF或png格式
# 輸出的dot文件可以使用graphvize軟件轉(zhuǎn)為PDF，graphvize安裝目錄中的bin目錄放入到環(huán)境變量的Path中
# 使用如下命令
#cmd
# dot -Tpdf  c:/tree.dot -o c:/tree.pdf
#下載地址：http://www.graphviz.org/Download_windows.php
#生成dot文件
with open("c:/tree.dot",'w') as f:
    f = tree.export_graphviz(clf, feature_names= vec.get_feature_names(),out_file= f)