# -*- coding: utf-8 -*-
import numpy as np
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import precision_recall_curve
from sklearn.metrics import classification_report
from sklearn.naive_bayes import BernoulliNB
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.cross_validation import train_test_split
import matplotlib.pyplot as plt
import pandas as pd
####knn最邻近算法####
inputfile = 'd:/data/sales_data.xls'
data = pd.read_excel(inputfile, index_col = u'序号') #导入数据
#数据是类别标签，要将它转换为数据
#用1来表示“好”、“是”、“高”这三个属性，用-1来表示“坏”、“否”、“低”
data[data == u'好'] = 1
data[data == u'是'] = 1
data[data == u'高'] = 1
data[data != 1] = -1
x = data.iloc[:,:3].as_matrix().astype(int)
y = data.iloc[:,3].as_matrix().astype(int)
#拆分训练数据与测试数据
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size = 0.2)
#训练KNN分类器
clf = KNeighborsClassifier(algorithm='kd_tree')
clf.fit(x_train, y_train)
#测试结果
answer = clf.predict(x_test)
print(x_test)
print(answer)
print(y_test)
print(np.mean( answer == y_test))
#准确率
precision, recall, thresholds = precision_recall_curve(y_train, clf.predict(x_train))
print(classification_report(y_test, answer, target_names = ['高', '低']))
####贝叶斯分类器####
#训练贝叶斯分类器
clf = BernoulliNB()
clf.fit(x_train,y_train)
#测试结果
answer = clf.predict(x_test)
print(x_test)
print(answer)
print(y_test)
print(np.mean( answer == y_test))
print(classification_report(y_test, answer, target_names = ['低', '高']))
####决策树####
from sklearn.tree import DecisionTreeClassifier as DTC
dtc = DTC(criterion='entropy') #建立决策树模型，基于信息熵
dtc.fit(x_train, y_train) #训练模型
#导入相关函数，可视化决策树。
#导出的结果是一个dot文件，需要安装Graphviz才能将它转换为pdf或png等格式。
from sklearn.tree import export_graphviz
from sklearn.externals.six import StringIO
with open("tree.dot", 'w') as f:
f = export_graphviz(dtc, out_file = f)
#测试结果
answer = dtc.predict(x_test)
print(x_test)
print(answer)
print(y_test)
print(np.mean( answer == y_test))
print(classification_report(y_test, answer, target_names = ['低', '高']))
####SVM####
from sklearn.svm import SVC
clf =SVC()
clf.fit(x_train, y_train)
#测试结果
answer = clf.predict(x_test)
print(x_test)
print(answer)
print(y_test)
print(np.mean( answer == y_test))
print(classification_report(y_test, answer, target_names = ['低', '高']))