对于这种，读进来日期不是日期格式，并且有时候一天有很多条数据，一天没有数据的情况，根据年份合并求star_rating这一栏的平均值。
 
 
读入数据并删除无关列
 
pcfr = pd.read_excel('microwave.xlsx')
df = pcfr[['review_date','star_rating']]
df.head(3)
 
会警告，无视之。
 
df['review_date'] = pd.to_datetime(df['review_date'])
 
日期格式的转化
 
# df.set_index('review_date')
# df.reset_index()
t = df.set_index('review_date')
t.head(3)
 
这一步结束后可以看到日期已经转换成功。
 
 利用groupby分组求年平均值
 
m = t.groupby(pd.Grouper(freq='Y')).mean()
print(m)
 

去掉某些列
 
delIx = strcmp(A, 'settings') | strcmp(A, 'exposure') | strcmp(A, 'equity'); 
A= A(~delIx);
 
 
 
合并去掉重复的
 
A= ([B, A(~ismember(A,B)) ]);
 
 
 
日期：1990，1,1 变成 19900101
 
datestr(datenum(1990,1,1),'yyyymmdd')
 
ans =
 
19900101
 
 
 
fgetl读取一行，取出fieldname类似excel里面的列名
 
然后送入textscan进行分析。
 
 
 
 
 
全部时间顺序
 
out.DATE = str2num(datestr(datenum(1990, 01, 01):datenum(date), 'yyyymmdd'));
 
截取部分
 
out.DATE = out.DATE(out.DATE >= settings.samplebegin); 
out.DATE = out.DATE(out.DATE <= settings.sampleend);
 
 
 
k线实体与前一天的close比值
 
(CLOSE(2:end,:) - OPEN(2:end,:)) ./ CLOSE(1:end-1,:) 
 
 
 
取最后n天的数据
 
CLOSE(end-n+1:end,:)
 
 
 
取出年 dateVec(1)=19911226
 
for j = floor(dateVec(1)/10^4):1:floor(dateVec(end)/10^4)
 
生成类似时间戳
 
dVec = datenum(([floor(dateVec/10^4) floor((mod(dateVec,10^4))/10^2) mod(dateVec,10^2)]));  
 
 
 
均值回归策略:
 
smashort与smalong比值是运动的，
 
lookback的k线数量
 
策略运动到504天，然后取出504的k线进入运算
 
策略运动到505天，然后取出504的k线进入策略计算 

 
目录：
 
效果图：
在线运行：


 网上都是合并列的，自己做了一个合并行的。需求是合并
相同数据的列。合并name1 name2 name3三个行
有空数据的行。
 
效果图：
 
 
<template>
  <div class="views-container">
    <el-table ref="tableFormatData" border :span-method="objectSpanMethod" :data="tableData" style="width: 100%">
      <el-table-column type="selection" width="55" align="center" />
      <el-table-column label="序号" width="55" align="center" prop="id" />
      <el-table-column label="订单编号" align="center">
        <template slot-scope="scope">
          <span>{{ scope.row.bjd }}</span>
        </template>
      </el-table-column>
      <el-table-column label="委托编号" align="center" width="80">
        <template slot-scope="scope">
          <span>{{ scope.row.ypbh }}</span>
        </template>
      </el-table-column>
      <el-table-column label="产品名称" align="center">
        <template slot-scope="scope">
          <span>{{ scope.row.name1 }}</span>
        </template>
      </el-table-column>
      <el-table-column label="规格型号" align="center">
        <template slot-scope="scope">
          <span>{{ scope.row.name2 }}</span>
        </template>
      </el-table-column>
      <el-table-column label="数量" align="center" width="60">
        <template slot-scope="scope">
          <span>{{ scope.row.name3 }}</span>
        </template>
      </el-table-column>
      <el-table-column label="单价" align="center" width="70">
        <template slot-scope="scope">
          <span>{{ scope.row.price }}</span>
        </template>
      </el-table-column>
      <el-table-column label="物流信息" align="center">
        <template slot-scope="scope">
          <p>顺丰快递</p>
          <p>JD0015972385147</p>
        </template>
      </el-table-column>
      <el-table-column label="报检日期" align="center" width="100">
        <template slot-scope="scope">
          <span>2020-04-27</span>
        </template>
      </el-table-column>
      <el-table-column label="状态" align="center" width="70">
        <template slot-scope="scope">
          <span>状态</span>
        </template>
      </el-table-column>
      <el-table-column label="操作" align="center" width="100">
        <template slot-scope="scope">
          <div class="operation-group">
            <el-button type="text">删除</el-button>
          </div>
        </template>
      </el-table-column>
    </el-table>
  </div>
</template>

<script>
export default {
  name: 'TaskList',
  data() {
    return {
      tableData: [{
        id: '1',
        bjd: 'BJ20200521001',
        ypbh: '000001',
        price: 20,
        name1: '隔音门',
        name2: '2018-11-14',
        name3: '14'
      },
      {
        id: '1',
        bjd: 'BJ20200521001',
        ypbh: '000002',
        price: 20,
        name1: '隔音门',
        name2: '',
        name3: ''
      },
      {
        id: '2',
        bjd: 'BJ20200521002',
        ypbh: '000004',
        price: 20,
        name1: '隔音门',
        name2: '14',
        name3: ''
      },
      {
        id: '2',
        bjd: 'BJ20200521002',
        ypbh: '000003',
        price: 40,
        name1: '隔音门',
        name2: '2018-11-14',
        name3: '14'
      }
      ],
      spanArr: [],
      trArr: []
    }
  },
  mounted() {
    // 合并表格
    var contactDot = 0
    const newTabdata = this.tableData
    this.tableData.forEach((item, index) => {
      console.log(item.name2.length, 'item.name2.length')
      // 合并行
      if (item.name2.length !== 0 && item.name3.length === 0) {
        this.trArr.push(1)
      } else if (item.name2.length === 0 && item.name3.length === 0) {
        this.trArr.push(2)
      } else {
        this.trArr.push(0)
      }
      // 合并列。根据bjd合并
      if (index === 0) {
        this.spanArr.push(1)
      } else {
        // 遍历需要合并的名字 bjd
        if (item.bjd === newTabdata[index - 1].bjd) {
          this.spanArr[contactDot] += 1
          this.spanArr.push(0)
        } else {
          contactDot = index
          this.spanArr.push(1)
        }
      }
    })
    console.log(this.trArr, 'ront')
  },
  methods: {
    objectSpanMethod({
      row,
      column,
      rowIndex,
      columnIndex
    }) {
      // console.log(rowIndex, columnIndex)
      var hbArr = null; var hbArr5 = null; var hbArr6 = null
      this.trArr.map((item, index) => {
        if (columnIndex === 4 && (rowIndex === index)) {
          if (item === 2) {
            hbArr = {
              rowspan: 1,
              colspan: 3
            }
          }
        }
        if (columnIndex === 5 && (rowIndex === index)) {
          if (item === 1) {
            hbArr5 = {
              rowspan: 1,
              colspan: 2
            }
          }
          if (item === 2) {
            hbArr5 = {
              rowspan: 0,
              colspan: 0
            }
          }
        }
        if (columnIndex === 6 && (rowIndex === index)) {
          if (item === 1) {
            hbArr6 = {
              rowspan: 0,
              colspan: 0
            }
          }
          if (item === 2) {
            hbArr6 = {
              rowspan: 0,
              colspan: 0
            }
          }
        }
      })
      // 返回合并列，合并列的空数据
      if (columnIndex === 4 && hbArr) {
        console.log(hbArr, 'hbArr')
        return hbArr
      }
      if (columnIndex === 5 && hbArr5) {
        console.log(hbArr5, 'hbArr5')
        return hbArr5
      }
      if (columnIndex === 6 && hbArr6) {
        console.log(hbArr6, 'hbArr6')
        return hbArr6
      }

      // columnIndex 0 就是第0列，后面相对改,columnIndex 0 就是第0列，后面相对改，合并第 0,1,2列相同数据，就是多选，订单编号，委托编号3列
      if (columnIndex === 0) {
        if (this.spanArr[rowIndex]) {
          return {
            rowspan: this.spanArr[rowIndex],
            colspan: 1
          }
        } else {
          return {
            rowspan: 0,
            colspan: 0
          }
        }
      }
      if (columnIndex === 1) {
        if (this.spanArr[rowIndex]) {
          return {
            rowspan: this.spanArr[rowIndex],
            colspan: 1
          }
        } else {
          return {
            rowspan: 0,
            colspan: 0
          }
        }
      }
      if (columnIndex === 2) {
        if (this.spanArr[rowIndex]) {
          return {
            rowspan: this.spanArr[rowIndex],
            colspan: 1
          }
        } else {
          return {
            rowspan: 0,
            colspan: 0
          }
        }
      }
    }
  }
}
</script>

 
在线运行：
 
在线运行

create table [Data_NewTrayGuide_Picture]
(
	[vpictureid] int IDENTITY(1,1) NOT NULL,
	[VID] int not null,
	[title] varchar(2000) not null
)
go

/**********插入测试数据*******************/
declare   @a int,@vid int   
  set @a=10   
  while   @a<=200   
  begin   
	set @vid=@a/4
     insert into [Data_NewTrayGuide_Picture] 
     values(@vid,'[thetitle]'+cast(@a as varchar(200)))       
     set   @a=@a+1  
  end
  go
  /***********正题开始******************/
  select * from [Data_NewTrayGuide_Picture]
  /**************************************************/
  SELECT * INTO #tt FROM Data_NewTrayGuide_Picture --吧数据添加到临时表里面去
go

/*********************************************************/
with Data_NewTrayGuide_Picture as
(
	select ORDERID=(select count(vid) FROM #tt T1 
	WHERE T1.vid=T2.vid AND T1.vpictureid<=T2.vpictureid), * FROM #tt T2
)
select vid ,MAX(CASE ORDERID WHEN 1 THEN title ELSE '' END) as vpicID1, 
			MAX(CASE ORDERID WHEN 2 THEN title ELSE '' END) as vpicID2,
			MAX(CASE ORDERID WHEN 3 THEN title ELSE '' END) as vpicID3
			from Data_NewTrayGuide_Picture
			group by vid
			order by vid
/***************************************************************/