#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time    : 2022/1/1 13:49
# @Author  : @linlianqin
# @Site    : 
# @File    : naivyBates.py
# @Software: PyCharm
# @description:
from sklearn.naive_bayes import GaussianNB
from sklearn.naive_bayes import BernoulliNB
from sklearn.naive_bayes import MultinomialNB
import csv
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import accuracy_score
from sklearn.metrics import confusion_matrix
from sklearn.metrics import classification_report

import joblib
from dataProcess import loaddatasets
from paths import abs_path
import numpy as np


# 处理数据
def loadDataSets(xlspath):
	datas, labels = loaddatasets(xlspath)
	labels_ = labels.reshape(len(labels), 1)
	dataSet = np.hstack((datas, labels_)).astype(int)

	return dataSet, datas, labels


# 训练
def train(traffic_feature, traffic_target):
	print('traffic_feature=', traffic_feature)
	print('traffic_target=', traffic_target)
	scaler = StandardScaler()  # 标准化转换
	scaler.fit(traffic_feature)  # 训练标准化对象
	traffic_feature = scaler.transform(traffic_feature)  # 转换数据集
	feature_train, feature_test, target_train, target_test = train_test_split(traffic_feature, traffic_target,
	                                                                          test_size=0.1, random_state=0)
	model = BernoulliNB()
	model.fit(feature_train, target_train)
	return model,feature_test,target_test


# 预测
def predict(model, feature_test):
	predict_results = model.predict(feature_test)
	return predict_results  # [1,2,3]


# 评估
def evalue(model, predict_labels, true_labels):
	acc = accuracy_score(predict_labels, true_labels)
	print("准确率：", acc)

	conf_mat = confusion_matrix(true_labels, predict_labels)
	print("混淆矩阵：", conf_mat)

	report = classification_report(true_labels, predict_labels)
	print("模型分析报告：", report)

	return acc, conf_mat, report


# 保存模型
def save_model(model, path):
	# 保存模型
	joblib.dump(model, path)


# 加载模型
def load_model(path):
	# 加载模型进行预测
	new_model = joblib.load(path)
	return new_model


if __name__ == '__main__':
	print("加载数据集......")
	xlsPath = abs_path + "\\data\\min_datas.xlsx"
	dataSet, datas, labels = loadDataSets(xlsPath)

	print("开始训练......")
	model,feature_test,target_test = train(datas, labels)

	print("测试模型，测试集")
	predict_labels = predict(model,feature_test)

	print("评估模型......")
	acc, conf_mat, report = evalue(model,predict_labels,target_test)

	print("保存模型")
	path = abs_path+"\\data\\naivyBates_%.2f.pkl"%acc
	save_model(model,path)

	print("调用模型进行预测")
	testVec = [[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0]] # 注意数据是二维的
	new_model = load_model(path)
	predict_results = new_model.predict(testVec)
	print("待测数据：",testVec)
	print("预测结果：",predict_results)