Question

我发现了link

这里的代码可以使用数据透视值和数据透视表来旋转数据表。我需要的是根据日期进行调整，并保留所有其他值列，如下所示：

Date Column1 Column2 Column3
2010   10      20      30
2011   40      50      60

在这里，我将列名称作为第一行及其在年份下的值：

          2010  2011 
Column1    10    40
Column2    20    50
Column3    30    60

这是我现在的代码：

DataTable Pivot(DataTable dt, DataColumn pivotColumn, DataColumn pivotValue) {
    // find primary key columns 
    //(i.e. everything but pivot column and pivot value)
    DataTable temp = dt.Copy();
    temp.Columns.Remove( pivotColumn.ColumnName );
    temp.Columns.Remove( pivotValue.ColumnName );
    string[] pkColumnNames = temp.Columns.Cast<DataColumn>()
        .Select( c => c.ColumnName )
        .ToArray();

    // prep results table
    DataTable result = temp.DefaultView.ToTable(true, pkColumnNames).Copy();
    result.PrimaryKey = result.Columns.Cast<DataColumn>().ToArray();
    dt.AsEnumerable()
        .Select(r => r[pivotColumn.ColumnName].ToString())
        .Distinct().ToList()
        .ForEach (c => result.Columns.Add(c, pivotColumn.DataType));

    // load it
    foreach( DataRow row in dt.Rows ) {
        // find row to update
        DataRow aggRow = result.Rows.Find(
            pkColumnNames
                .Select( c => row[c] )
                .ToArray() );
        // the aggregate used here is LATEST 
        // adjust the next line if you want (SUM, MAX, etc...)
        aggRow[row[pivotColumn.ColumnName].ToString()] = row[pivotValue.ColumnName];
    }

    return result;
}

如何使用多个列值实现此结果？

Answer 1

我认为你需要采取略微不同的方法，因为你的方式略有不同。这是一种仅使用表格和pivotColumnName的方法。使用输入表计算出其他所有内容。

DataTable Pivot(DataTable table, string pivotColumnName)
{
    // TODO make sure the table contains at least two columns

     // get the data type of the first value column
    var dataType = table.Columns[1].DataType;

    // create a pivoted table, and add the first column
    var pivotedTable = new DataTable();
    pivotedTable.Columns.Add("Row Name", typeof(string));

    // determine the names of the remaining columns of the pivoted table
    var additionalColumnNames = table
        .AsEnumerable()
        .Select(x => x[pivotColumnName].ToString());

    // add the remaining columns to the pivoted table
    foreach (var columnName in additionalColumnNames)
        pivotedTable.Columns.Add(columnName, dataType);

    // determine the row names for the pivoted table
    var rowNames = table.Columns
        .Cast<DataColumn>()
        .Select(x => x.ColumnName)
        .Where(x => x != pivotColumnName);

    // fill in the pivoted data
    foreach (var rowName in rowNames)
    {
        // get the value data from the appropriate column of the input table
        var pivotedData = table
            .AsEnumerable()
            .Select(x => x[rowName]);

        // make the rowName the first value
        var data = new object[] { rowName }
            .Concat(pivotedData)
            .ToArray();

        // add the row
        pivotedTable.Rows.Add(data);
    }

    return pivotedTable;
}

以下是我用来测试的内容：

var dateColumn = new DataColumn("Date", typeof(int));
var column1 = new DataColumn("Column1", typeof(int));
var column2 = new DataColumn("Column2", typeof(int));
var column3 = new DataColumn("Column3", typeof(int));
var table = new DataTable();
table.Columns.Add(dateColumn);
table.Columns.Add(column1);
table.Columns.Add(column2);
table.Columns.Add(column3);
table.Rows.Add(new object[] { 2010, 10, 20, 30 });
table.Rows.Add(new object[] { 2011, 40, 50, 60 });
var pivotedTable = Pivot(table, "Date");

使用许多值列透视数据表c＃

1 个答案: