dalian-university/秀水分区/data_analyse.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# time: 2023/4/12 18:57
# file: data_analyse.py
# author: David
# company: shenyang JY
import sys
import numpy as np
np.random.seed(42)
import matplotlib.pyplot as plt
import pandas as pd
from calculate import calculate_acc


class data_analyse(object):
    def __init__(self, opt, logger):
        self.opt = opt
        self.logger = logger

    def dq_acc(self):
        excel_data_path = self.opt.excel_data_path
        data_format = self.opt.data_format
        dq_path = excel_data_path + data_format["dq"]
        dq_columns = ['C_TIME', 'C_ABLE_VALUE']
        dq = pd.read_csv(dq_path, usecols=dq_columns)
        dq['C_TIME'] = pd.to_datetime(dq['C_TIME'], format='%Y-%m-%d %H:%M:%S')
        return dq

    def cdq_acc(self):
        excel_data_path = self.opt.excel_data_path
        data_format = self.opt.data_format
        dq_path = excel_data_path + data_format["cdq"]
        dq_columns = ['C_TIME', 'C_ABLE_VALUE']
        cdq = pd.read_csv(dq_path, usecols=dq_columns)
        cdq['C_TIME'] = pd.to_datetime(cdq['C_TIME'], format='%Y-%m-%d %H:%M:%S')
        return cdq

    def calculate_acc(self, label_data, predict_data):
        loss = np.sum((label_data - predict_data) ** 2) / len(label_data)  # mse
        loss_sqrt = np.sqrt(loss)  # rmse
        return loss_sqrt

    def predict_acc(self, predict_data, dfs, predict_all=False):
        if predict_all is True:
            # predict_data = predict_data * self.opt.std['C_VALUE'] + self.opt.mean['C_VALUE']
            predict_data = predict_data * self.opt.std['sum_power'] + self.opt.mean['sum_power']
        else:
            predict_data0 = predict_data[0] * self.opt.std['col1_power'] + self.opt.mean['col1_power']
            predict_data1 = predict_data[1] * self.opt.std['col2_power'] + self.opt.mean['col2_power']
            predict_data = predict_data0 + predict_data1
        dfs1 = []
        dfs2 = []
        for i, df in enumerate(dfs):
            df["forecastAbleValue"] = predict_data[i]
            dfs1.append(df.iloc[0])  # 第1个点
            dfs2.append(df.iloc[-1])
            # if df.iloc[-1, -1] < 0:
            #     print("预测中有一个负值，为：", df.iloc[-1, -1])
            # else:
            #     print("预测结果为：", df.iloc[-1, -1])
            #     dfs1.append(df.iloc[0])
            # if df.iloc[-1, -1] < 0:
            #     print("预测中有一个负值，为：", df.iloc[-1, -1])
            # else:
            #     print("预测结果为：", df.iloc[-1, -1])
            #     dfs2.append(df.iloc[-1])
            # dfs[i] = df.iloc[self.opt.predict_point]  # 第16个点
        df1 = pd.concat(dfs1, axis=1).T
        df2 = pd.concat(dfs2, axis=1).T
        # df = pd.concat(dfs, axis=1).T
        # df1 = df.drop(['label'], axis=1)
        # df1 = df1.iloc[15:, :]
        # df2 = df2.iloc[:-15, :]
        # fig, ax = plt.subplots()
        # ax.plot(df1["C_TIME"], df1["forecastAbleValue"], color='b')
        # ax.plot(df2["C_TIME"], df2["forecastAbleValue"], color='r')
        # ax.plot(df2["C_TIME"], df2["C_VALUE"], color='y')
        # plt.show()
        # rmse = self.calculate_acc(label_data=df['realValue'], predict_data=df['forecastAbleValue'])
        # df1.to_csv('./figure/fenqu.csv')
        self.logger.info("1 新模型预测准确率是： {} ".format('接口') + str(self.cal_acc(df1)))
        self.logger.info("16 新模型预测准确率是： {} ".format('接口') + str(self.cal_acc(df2)))
        # self.logger.info("新模型预测rmse是： {} ".format('公式') + str(rmse))
        # self.predict_draw(df1)
        # self.logger.info("1 新模型预测准确率是： {} ".format('接口') + str(self.calculate_acc(df1['realValue'], df1['forecastAbleValue'])))
        # self.logger.info("16 新模型预测准确率是： {} ".format('接口') + str(self.calculate_acc(df2['realValue'], df2['forecastAbleValue'])))

    def cal_acc(self, df):
        df.rename(columns={'C_VALUE': 'realValue'}, inplace=True)
        df['ableValue'] = df['realValue']
        acc = calculate_acc(df, self.opt)
        return acc

    def calculate_acc(self, label_data, predict_data):
        loss = np.sum((label_data - predict_data) ** 2) / len(label_data)  # mse
        loss_sqrt = np.sqrt(loss)  # rmse
        loss_acc = 1 - loss_sqrt / self.opt.cap
        return loss_acc

    def predict_draw(self, df):
        df.realValue.plot()
        df.forecastAbleValue.plot()
        plt.show()


if __name__ == '__main__':
    pass