#构建测试数据
df = pd.DataFrame({
	'品类': ['苹果', '橘子', '香蕉', '橘子', '香蕉', '橘子'],
	'日期': ['2023-10-21', '2023-10-21', '2023-10-21', '2023-10-22', '2023-10-23','2023-10-23'],
	'销量': [13, 24, 25, 27, 19, 21]
	},
	columns=['品类', '日期', '销量']
)
print(df)
#日期列转行 index 行索引，columns 要转为列名的行，values 要转为列值的行
df2 = pd.pivot_table(df, index=['品类'], columns=['日期'], values=['销量'])
df2.fillna(0, inplace=True)
print(df2)

  #为订单数据集增加以后一列：微信支付的金额
  df['pay_wx'] = df.apply(lambda x: (x.pay if x.chan == 'wx' else 0), axis=1)

  df_wx = df[df['chan'] == 'wx']
  df_wx.rename(columns={'pay': 'pay_wx'}, inplace=True)
  df['pay_wx'] = df_wx['pay_wx']
  df.fillna({'pay_wx': 0}, inplace=True)

  #合并某项分类依据
  df['color'] = df.apply(lambda x: ('红' if x.color == '粉红' else x.color), axis=1)
  #使用merge优化
  df_h = df[df['color'] == '粉红'][['id', 'color']]
  df_h['color'] = '红'
  df_nh = df[df['color'] != '粉红'][['id', 'color']]
  df_hs = pd.concat([df_h, df_nh])
  df_hs.rename(columns={'color':'color2'}, inplace=True)
  df = pd.merge(df, df_h, on='id', how='left')
  df['color'] = df['color2']

import matplotlib.pyplot as plt
from matplotlib.font_manager import FontProperties
# figsize 输出尺寸，默认1000×300
# savePath 保存路径
# title 折线图上显示的标题
# rightColumns 放在右侧的字段
# leftLabel  rightLabel  左右标签
def buildChartLine(dataDf, savePath, xInd='日期', figsize=(10, 3), title='', rightColumns=[], leftLabel='', rightLabel='', colors=[]):
    if isLocalServ(): # macos本地的中文配置
        plt.rcParams['font.sans-serif'] = ['SimHei']
        plt.rcParams['axes.unicode_minus'] = False
    else: # Linux的中文配置
        tfont = FontProperties(fname=r"/usr/share/fonts/simhei.ttf", size=14)
    dataDf.sort_values(xInd, inplace=True) # 确保顺序
    dataDf = dataDf.copy()
    dataDf.set_index(xInd, inplace=True) # 确保索引x轴字段名
    if len(colors) == 0:
        #TODO: 自定义颜色策略
        print('no color')
    plt.figure()
    if isLocalServ(): # macos本地的中文配置，和下面else里的，主要还是中文配置的差异，其他相同
        if len(rightColumns) > 0:
            ax = dataDf.plot(secondary_y=rightColumns, figsize=figsize, color=colors, kind='line')
            ax.right_ax.set_ylabel(rightLabel)
        else:
            ax = dataDf.plot(figsize=figsize, color=colors, kind='line')
        ax.set_ylabel(leftLabel)
        if len(title) > 0: ax.set(title=title)
        ax.legend(loc=2)
        plt.legend(loc=1)
    else:
        if len(rightColumns) > 0:
            ax = dataDf.plot(secondary_y=rightColumns, figsize=figsize, color=colors, kind='line')
            ax.right_ax.set_ylabel(rightLabel, fontproperties=tfont)
        else:
            ax = dataDf.plot(figsize=figsize, color=colors, kind='line')
        ax.set_ylabel(leftLabel, fontproperties=tfont)
        if len(title) > 0: plt.title(title, fontproperties=tfont)
        ax.legend(loc=2, prop=tfont)
        plt.legend(loc=1, prop=tfont)
    # plt.show() #直接显示折线图
    plt.savefig(savePath)

dataDf = pd.DataFrame({
	'日期': ['2021-11-01', '2021-11-02', '2021-11-04',  '2021-11-06',  '2021-11-03',  '2021-11-05'],
	'风力': [3, 4, 5, 7, 9,	1],
	'最高温度': [13, 24, 25, 27, 19, 21],
	'最低温度': [1, 4, 5, 7, 9, 2]},
	columns=['日期', '风力', '最高温度', '最低温度'])
buildChartLine(dataDf, 'dayInfo', xInd='日期', figsize=(10, 3), title='气象记录', rightColumns=['风力'], leftLabel='温度（℃）', rightLabel='风力（级）', colors=['#fb0320', '#e50ce2', '#770ce5'])

#构建测试数据
df1 = pd.DataFrame({
	'uid': ['小A',  		'小B',  		'小A',  		'小A',  		'小A',  		'小B',  		  '小B',  	     '小A',  		'小B',  	   '小A'],
	'day': ['2021-09-25',  	'2021-09-25',  '2021-09-25',  	'2021-09-26',  '2021-09-26',  '2021-09-26',  	'2021-09-26', '2021-09-27',  '2021-09-25',  '2021-09-25'],
	'sub': ['语文', 		'语文', 		'数学', 		'语文', 		 '数学', 		  '语文',           '数学',        '语文',         '英语',        '英语'],
	'val': [95,             95,            74,             87,             65,             77,              66,            32 ,           55,            90]}, 
		columns=['uid', 'day', 'sub', 'val'])
sortList = ['数学', '语文', '英语']  #排序依据数组，依据sub，科目
# df1.reset_index(inplace=True) #视数据实际情况，排序字段不能是索引
df1['sub'] = df1['sub'].astype('category') #重新设置字段类型，为排序做准备
df1['sub'].cat.set_categories(sortList, inplace=True) #设置排序依据
df1.sort_values('sub', inplace=True) #排序操作
print(df1)

import pandas as pd
from openpyxl import Workbook
from openpyxl.reader.excel import load_workbook
 
#数据暂存容器
xlsDatas = {}
#保存路径
savePath = '.../test-excel.xlsx'
#获取数据
df1 = pd.DataFrame(
    {
        '日期': ['2021-01-01', '2021-01-02', '2021-01-03', '2021-01-04'], 
        '订单数': [1234, 1321, 3456, 6543],
        '用户数': [234, 787, 987, 545]},
    columns=['日期', '订单数', '用户数']
)
xlsDatas['概况'] = df1  #子表名-子表数据
df2 = pd.DataFrame(
    [
        [56335, '兰州拉面', 15.0, '2021-01-01 12:12:12'], 
        [45135, '山西削面', 13.0, '2021-01-01 12:16:12'], 
        [16335, '兰州拉面', 15.0, '2021-01-01 12:32:12']
    ],
    index=list('123'),
    columns=['用户id', '产品名称', '消费金额', '消费日期']
)
xlsDatas['订单列表'] = df2
#使用Workbook保存数据
book0 = Workbook()
book0.save(savePath) #先保存一个空excel文件，然后往里塞子表，实际会有一个空的Sheet子表
book = load_workbook(savePath)
writer = pd.ExcelWriter(savePath, engine='openpyxl') #选择openpyxl引擎
writer.book = book
for k in xlsDatas :
    xlsDatas[k].to_excel(writer, sheet_name=k)
writer.save()

  orderDf = pd.read_csv('path...')  #订单数据
  nodePlaces = ... #查库，返回 点位编号 - 场景类型 字典数据
  orderDf['place'] = orderDf.apply(lambda x: nodePlaces[x.nodeId] if x.nodeId in nodePlaces else 0, axis = 1) //#补齐场景类型字段place
  print(orderDf.colums)  #查库补齐效果
  print(orderDf.tail(5))

  orderDf = pd.read_csv('path...')  #订单数据
  nodePlaces = ... #查库，返回 点位编号 - 场景类型 字典数据
  nodeDf = pd.DataFrame(list(nodePlaces()), columns=['nodeId', 'place']) #转成datafram
  orderDf = pd.merge(orderDf, nodeDf, on='nodeId', how='left') //#补齐场景类型字段place
  print(orderDf.colums)  #查库补齐效果
  print(orderDf.tail(5))