import os
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.graph_objects as go
from behavior.analystExcludeRatedPower import AnalystExcludeRatedPower
from utils.directoryUtil import DirectoryUtil as dir
from algorithmContract.confBusiness import *


class TSRTrendAnalyst(AnalystExcludeRatedPower):
    """
    风电机组叶尖速比时序分析
    """

    def typeAnalyst(self):
        return "tsr_trend"

    def turbinesAnalysis(self, dataFrameMerge, outputAnalysisDir, confData: ConfBusiness):
        self.drawTSRTrend(dataFrameMerge, outputAnalysisDir, confData)

    def drawTSRTrend(self,dataFrameMerge:pd.DataFrame, outputAnalysisDir, confData: ConfBusiness):        
        # 检查所需列是否存在
        required_columns = {Field_TSR, Field_YearMonthDay}
        if not required_columns.issubset(dataFrameMerge.columns):
            raise ValueError(f"DataFrame缺少必要的列。需要的列有: {required_columns}")
        
        # 按设备名分组数据
        grouped = dataFrameMerge.groupby(Field_NameOfTurbine)

        for name, group in grouped:
            # 计算四分位数和IQR
            Q1 = group[Field_TSR].quantile(0.15)
            Q3 = group[Field_TSR].quantile(0.90)
            IQR = Q3 - Q1
            # 定义离群值的范围
            lower_bound = Q1 - 1.5 * IQR
            upper_bound = Q3 + 1.5 * IQR

            # 筛选掉离群值
            filtered_group = group[(group[Field_TSR] >= lower_bound) & (group[Field_TSR] <= upper_bound)]

            # 创建箱线图
            fig = go.Figure()

            fig.add_trace(go.Box(
                x=filtered_group[Field_YearMonthDay],  # 设置x轴数据为日期
                y=filtered_group[Field_TSR],  # 设置y轴数据为风能利用系数
                # boxpoints='outliers',  # 显示异常值（偏离值），不显示数据的所有点（只显示异常值）
                boxpoints=False,  # 不显示偏离值
                marker=dict(color='lightgoldenrodyellow', size=1),  # 设置偏离值的颜色和大小
                line=dict(color='lightgray', width=2),  # 设置箱线和须线的颜色为灰色，粗细为2
                fillcolor='rgba(200, 200, 200, 0.5)',  # 设置箱体的填充颜色和透明度
                name='TSR'  # 图例名称
            ))

            # 对于每个箱线图的中位数，绘制一个蓝色点
            medians = filtered_group.groupby(filtered_group[Field_YearMonthDay])[Field_TSR].median()
            fig.add_trace(go.Scatter(
                x=medians.index,
                y=medians.values,
                mode='markers',
                marker=dict(color='orange', size=3),
                name='Median TSR'  # 中位数标记的图例名称
            ))

            # 设置图表的标题和轴标签
            fig.update_layout(
                title={
                    'text': f'TSR Trend Turbine Name {name}',
                    'x':0.5,
                },
                xaxis_title='time',
                yaxis_title='TSR',                
                xaxis=dict(
                    tickmode='auto',  # 自动设置x轴刻度，以适应日期数据
                    tickformat='%Y-%m-%d',  # 设置x轴时间格式
                    showgrid=True,  # 显示网格线
                    gridcolor='lightgray',  # setting y-axis gridline color to black
                    tickangle=-45,
                    linecolor='black',  # 设置y轴坐标系线颜色为黑色
                    ticklen=5,  # 设置刻度线的长度
                ),
                yaxis=dict(
                    dtick=confData.graphSets["tsr"]["step"] if not self.common.isNone(
                        confData.graphSets["tsr"]["step"]) else 5,  # 设置y轴刻度间隔为0.1
                    range=[confData.graphSets["tsr"]["min"] if not self.common.isNone(
                        confData.graphSets["tsr"]["min"]) else 0, confData.graphSets["tsr"]["max"] if not self.common.isNone(confData.graphSets["tsr"]["max"]) else 20],  # 设置y轴的范围从0到1
                    showgrid=True,  # 显示网格线
                    gridcolor='lightgray',  # setting y-axis gridline color to black
                    linecolor='black',  # 设置y轴坐标系线颜色为黑色
                    ticklen=5,  # 设置刻度线的长度
                ),  
                paper_bgcolor='white',  # 设置纸张背景颜色为白色  
                plot_bgcolor='white',  # 设置图表背景颜色为白色  
                margin=dict(t=50, b=10)  # t为顶部（top）间距，b为底部（bottom）间距
            )
            
            # 保存图像
            output_file = os.path.join(outputAnalysisDir, f"{name}.png")
            fig.write_image(output_file, scale=2)