We Keep Coding

Row already belongs to another table error when trying to add rows?

I tried this solution below:
This Row already belongs to another table error when trying to add rows?
I have a Datatable that contains 597 Columns and 20 Rows and are trying to export the data to excel. However, Excel has a maximum column count 256 and so I need to divide the source data into 3 datatables to make the export work.
Below is the code I have written.
var dtmasterdata = data.Tables[name];
for (int j = 1; j < datatableNumberCount; j++)
{
DataTable dt2 = new DataTable();
dt2.TableName = "Master_" + j;
dt2 = dtmasterdata.Copy();
foreach (DataColumn col in dtmasterdata.Columns)
{
DataColumn dtcol = new DataColumn();
dtcol = col;
dt2.Columns.Add(dtcol.ColumnName, dtcol.DataType);
}
for (int k = 0; k < dtmasterdata.Rows.Count; k++)
{
DataRow dr = dt2.NewRow();
dr = dtmasterdata.Rows[k];
dt2.ImportRow(dtmasterdata.Rows[k]);
//dt2.Rows.Add(dr.ItemArray);
}
After that I need to delete few columns like below and I want to create 3 datatables
foreach (DataColumn col in dtmasterdata.Columns)
{
if (j == 1)
{
// condition 1
if (col.Ordinal >= 255)
{
dt2.Columns.RemoveAt(col.Ordinal);
}
}
if (j == 2)
{
// condition 2.
if (col.Ordinal < 255 || col.Ordinal >= 510)
{
dt2.Columns.RemoveAt(col.Ordinal);
}
}
if (j == 3)
{
// condition 3.
if (col.Ordinal <= 510 || col.Ordinal >= 765)
{
dt2.Columns.Add(col);
}
}
}
int worksheetNumber = 1;
string worksheetNameWithNumber = "Master Data";
if (worksheetNumber > 1)
worksheetNameWithNumber = String.Format("{0}_{1}", ws1, worksheetNumber.ToString());
Infragistics.Excel.Worksheet worksheet = wb.Worksheets.Add(worksheetNameWithNumber);
Infragistics.WebUI.UltraWebGrid.UltraWebGrid masterData1 = new Infragistics.WebUI.UltraWebGrid.UltraWebGrid("masterDataGrid");
masterData1.Browser = Infragistics.WebUI.UltraWebGrid.BrowserLevel.UpLevel;
masterData1.DataSource = dt2;
masterData1.DataMember = "Master_" + j;
masterData1.DisplayLayout.HeaderStyleDefault.Font.Bold = true;
masterData1.DisplayLayout.HeaderStyleDefault.Font.Name = "Arial";
masterData1.DisplayLayout.HeaderStyleDefault.Font.Size = FontUnit.Parse("10px");
masterData1.DisplayLayout.HeaderStyleDefault.BackColor = System.Drawing.Color.LightGray;
masterData1.DisplayLayout.RowStyleDefault.Font.Name = "Arial";
masterData1.DisplayLayout.RowStyleDefault.Font.Size = FontUnit.Parse("10px");
Infragistics.WebUI.UltraWebGrid.UltraGridBand masterBand1 = new Infragistics.WebUI.UltraWebGrid.UltraGridBand();
masterData1.Bands.Add(masterBand1);
dgResults.Controls.Add(masterData1);
masterData1.DataBind();
wb.ActiveWorksheet = worksheet;
this.ugWebGridExporter.Export(masterData1, worksheet);
worksheetNumber++;

Your error is because you are trying to add a column to a datatable that already belongs to your source datatable.
dt2.Columns.Add(col);
You can't just iterate through the columns of a datatable and add them to another.
I've a solution to this, which involves cloning the source data and removing what you don't need.
1st, make 3 clones of the datatables you need. Below is an example with me creating my own source table with 596 columns. Notice that clone only takes the data table structure, no data!
var source597ColsTable = new DataTable("Source");
for (var i = 0; i <= 596; i++)
{
source597ColsTable.Columns.Add(new DataColumn("Column" + i , typeof(string)));
}
DataRow newRow = source597ColsTable.NewRow();
source597ColsTable.Rows.Add(newRow);
var cols0To199Table = source597ColsTable.Clone();
var cols200To399Table = source597ColsTable.Clone();
var cols400To596Table = source597ColsTable.Clone();
Next copy all the rows from the source table into the clones. The below is a simple function to do so.
private DataTable CopyRowsFromSource(DataTable sourceTable, DataTable destinationTable)
{
foreach (DataRow row in sourceTable.Rows)
{
destinationTable.Rows.Add(row.ItemArray);
}
return destinationTable;
}
Then call this function for each of your tables.
cols0To199Table = CopyRowsFromSource(source597ColsTable, cols0To199Table);
cols200To399Table = CopyRowsFromSource(source597ColsTable, cols200To399Table);
cols400To596Table = CopyRowsFromSource(source597ColsTable, cols400To596Table);
Finally, remove all the columns from the datatables to give you your split.
private DataTable RemoveColumns(DataTable table, int startCol, int endCol)
{
var colsToRemove = new List<DataColumn>();
for (var colCount = startCol; colCount <= endCol; colCount++)
{
colsToRemove.Add(table.Columns[colCount]);
}
foreach (DataColumn col in colsToRemove)
{
table.Columns.Remove(col);
}
return table;
}
Then call.. again for each cloned table.
cols0To199Table = RemoveColumns(cols0To199Table, 200, 596);
cols200To399Table = RemoveColumns(cols200To399Table, 0, 199);
cols200To399Table = RemoveColumns(cols200To399Table, 200, 396);
cols400To596Table = RemoveColumns(cols400To596Table, 0, 399);
After running this, you will have 3 datatables, columns 0-199, 200-399 and 400-596.
Hope that helps.

I am not sure to have really understood all of your code, but to copy a subset of columns to another datatable there is a very simple method in the DataView class named ToTable where you can list the columns you want in the new table. As added bonus, this method copies also the data in the 20 rows of your original table.
So the only difficult is to list these columns to the method.
You can proceed in this way using linq over the DataColumn collection
string[] firstCols = dtmasterdata.Columns
.Cast<DataColumn>()
.Take(255)
.Select(x => x.ColumnName).ToArray();
string[] secondCols = dtmasterdata.Columns
.Cast<DataColumn>()
.Skip(255)
.Take(255)
.Select(x => x.ColumnName).ToArray();
string[] thirdCols = dtmasterdata.Columns
.Cast<DataColumn>()
.Skip(510)
.Select(x => x.ColumnName).ToArray();
DataTable t1 = dtmasterdata.DefaultView.ToTable("Master_1", false, firstCols);
DataTable t2 = dtmasterdata.DefaultView.ToTable("Master_2", false, secondCols);
DataTable t3 = dtmasterdata.DefaultView.ToTable("Master_3", false, thirdCols);

How to keep all rows different in two DataTable?

I have two DataTable, dt1, dt2. I write a class to compare two DataTable and get rows different.
table dt1:
Col 1 Col 2 Col 3
A 8 #
B 21 ()
table dt2:
Col 1 Col 2 Col 3
A 8 ^%^%^%
C 827 _++)
It show line different like:
B 21 ()
I post my function to compare:
public static DataTable CompareDataTables(DataTable first, DataTable second)
{
first.TableName = "FirstTable";
second.TableName = "SecondTable";
//Create Empty Table
DataTable table = new DataTable("Difference");
try
{
//Must use a Dataset to make use of a DataRelation object
using (DataSet ds = new DataSet())
{
//Add tables
ds.Tables.AddRange(new DataTable[] { first.Copy(), second.Copy() });
//Get Columns for DataRelation
DataColumn[] firstcolumns = new DataColumn[ds.Tables[0].Columns.Count];
for (int i = 0; i < 2; i++)
{
firstcolumns[i] = ds.Tables[0].Columns[i];
}
DataColumn[] secondcolumns = new DataColumn[ds.Tables[1].Columns.Count];
for (int i = 0; i < 2; i++)
{
secondcolumns[i] = ds.Tables[1].Columns[i];
}
//Create DataRelation
DataRelation r = new DataRelation(string.Empty, firstcolumns, secondcolumns, false);
ds.Relations.Add(r);
//Create columns for return table
for (int i = 0; i < first.Columns.Count; i++)
{
table.Columns.Add(first.Columns[i].ColumnName, first.Columns[i].DataType);
}
//If First Row not in Second, Add to return table.
table.BeginLoadData();
foreach (DataRow parentrow in ds.Tables[0].Rows)
{
DataRow[] childrows = parentrow.GetChildRows(r);
if (childrows == null || childrows.Length == 0)
table.LoadDataRow(parentrow.ItemArray, true);
}
table.EndLoadData();
}
}
catch (Exception ex) { }
return table;
}
tableDifferent will contain all rows different.
DataTable tableDifferent;
tableDifferent = CompareDataTables(dt1, dt2);
I tried with code:
var rowsToDelete = from r1 in dt1.AsEnumerable()
join r2 in tableDifferent.AsEnumerable()
on r1.Field<String>("StudentID") equals r2.Field<String>("StudentID")
select r1;
Now, I want to remove all rows in dt1 and dt2, only keep all rows in tableDifferent.
Have any method to do this?

If the only problem left is to clear the other data tables, you can use the DataTable.Clear Method ()
All rows in all tables are removed. An exception is generated if the
table has any enforced child relations that would cause child rows to
be orphaned.
If the DataSet is bound to an XmlDataDocument, calling DataSet.Clear
or DataTable.Clear raises the NotSupportedException. To avoid this
situation, traverse each table, removing each row one at a time. When
you use NewRow to create new rows, the rows must be processed before
you call Clear
In your case:
dt1.Clear();
dt2.Clear();

Converting a List into Datatable

Assume that I have a list like PL = { P1, 0, 10, P2, 5, 20 } and I need to convert it a datatable like
ProcessName ArrivalTime CpuTime
P1 0 10
P2 5 20
The number of process (row count) is dynamic. And I have tried sth like this:
protected DataTable CreateDataTable()
{
int j = 0;
List<string> PL = CreateProcessList();
DataTable DT = new DataTable();
for (int i = 0; i < PL.Count - 2; i += 3)
{
DataRow ProcessRow = DT.NewRow();
DT.Rows[j][0] = PL[i].ToString();
DT.Rows[j][1] = Convert.ToInt32(PL[i + 1]);
DT.Rows[j][2] = Convert.ToInt32(PL[i + 2]);
j++;
}
DT.Columns.Add("Header", typeof(string));
DT.Columns[0].ColumnName = "ProcessName";
DT.Columns[1].ColumnName = "ArrivalTime";
DT.Columns[2].ColumnName = "CpuTime";
return DT;
}
It does not work (says that there is no row at position 0). Thanks for any idea.
Working Code After Editions:
protected DataTable CreateDataTable()
{
List<string> PL = CreateProcessList();
DataTable DT = new DataTable();
DT.Columns.Add("ProcessName", typeof(string));
DT.Columns.Add("ArrivalTime", typeof(int));
DT.Columns.Add("CpuTime", typeof(int));
for (int i = 0; i < PL.Count - 2; i += 3)
{
DataRow ProcessRow = DT.NewRow();
ProcessRow[0] = PL[i].ToString();
ProcessRow[1] = Convert.ToInt32(PL[i + 1]);
ProcessRow[2] = Convert.ToInt32(PL[i + 2]);
DT.Rows.Add(ProcessRow);
}
return DT;
}

To create a datatable in the way you have described, you need to follow a different way.
Create a datatable object
Add Columns to the datatable object by using the Add() method
use the datatable objects NewRow() method to get a DataRow object with the same schema as your datatable
populate the columns of this DataRow with the desired values
Add this DataRow to the Rows collection of your datatable object by using the Add() method
repeat step 3 to 6 until your list reaches the end.

On the second iteration of your for loop i is 3, so you're getting the 4th row of your table (at this point in time your table has 2 row). You're then getting the 4th, 5th, and 6th columns of that row (your table has 0 columns, as you haven't added any columns yet) to set their value. The corresponding index out of range errors should be telling you exactly what's wrong here.
Don't access the i-th row from the table. Just use ProcessRow to access the row; it's right there in a variable for you. Don't access the i-th column, access the 1st, 2nd, and 3rd (and add the columns before you try to populate them.

Something like this should work. Keep adding to curRow until the current iteration starts with 'P'. When it does start with 'P', add the currentRow to the data table and start a new row.
DataTable dataTable;
DataRow curRow;
... add columns to dataTable
for (var i = 0; i < PL.Count; i++) {
if (PL[i].ToString().StartsWith("P")) {
if (curRow != null)
dataTable.Rows.Add(curRow);
curRow = dataTable.NewRow();
}
... add PL[i] to curRow
}
There's a few little issues with this but they can be fixed pretty easily. A check to make sure curRow is not null before adding to data table... that sorta thing.

how to compare 2 datatable and get unique records in 3rd datatable?

I have 2 datatable let say dtData1 and dtData2. I have records in both the datatable I want to compare both the data table and want to create a new datatable let say dtData3 with uniqe records.
suppose: dtData1 has 10 records, and dtData2 has 50 records, but what ever records are there in dtData2 in that 7 records are same as dtData1. So here I want unique records in dtData3, means I want only 43 records in dtData3.
we don't know in datatable where we have the duplicate, and its possible that we will not have duplicates or its also possible that we will have all duplicate records.
So in dtData3 I need unique records
Some one please help me.

var dtData3 = dtData2.AsEnumerable().Except(dtData1.AsEnumerable(), DataRowComparer.Default);

Use this.. Probably it will help you.
Suppose you have two data table
DataTable dt1 = new DataTable();
dt1.Columns.Add("Name");
dt1.Columns.Add("ADD");
DataRow drow;
for (int i = 0; i < 10; i++)
{
drow = dt1.NewRow();
drow[0] = "NameA" + 1;
drow[1] = "Add" + 1;
dt1.Rows.Add();
}
DataTable dt2 = new DataTable();
dt2.Columns.Add("Name");
dt2.Columns.Add("ADD");
DataRow drow1;
for (int i = 0; i < 11; i++)
{
drow1 = dt2.NewRow();
drow1[0] = "Name" + 1;
drow1[1] = "Add" + 1;
dt2.Rows.Add();
}
Now To solve your problem Call :-
DataTable d3 = CompareTwoDataTable(dt1, dt2);
The method is something like this;--
public static DataTable CompareTwoDataTable(DataTable dt1, DataTable dt2)
{
dt1.Merge(dt2);
DataTable d3 = dt2.GetChanges();
return d3;
}

Then where the need to compare dtData1 and dtData2? Instead you can copy the contents from dtData2 to dtData3 starting from index7.

First Create Array variables to save unique coloumn values for datatable3. Use Foreach loop with second gridview rows. If any match then dnt save it in a array value, if dnt match then save it in arrays. and display it by attaching with third gridview.......
e.g
string[] name = new string[4];
int i=0,j=0;
foreach(GridViewRows gv in GridView1.rows)
{
if(gv.Cells[0].Text == GridView2.Rows[i].Cells[0].Text)' //if match
{
// dnt save
}
else' //if dnt match save in array for further use
{
name[j] = gv.Cells[0].Text;
j= j++;
}
i=i++;
}
After Saving unique values in Array "name"...Bind it in Third Gridview
During DataBound of third Gridview add this method...
Private void GridView3_RowDataBound(object sender,EventArgs e)
{
if(e.Row.RowState == DataControlState.DataRow)
{
foreach(string nm in name)
{
e.Rows.Cells.Add(name);
}
}
}

public DataTable CompareTwoDataTable(DataTable dtOriginalTable, DataTable dtNewTable, ArrayList columnNames)
{
DataTable filterTable = new DataTable();
filterTable = dtNewTable.Copy();
string filterCriterial;
if (columnNames.Count > 0)
{
for (int iNewTableRowCount = 0; iNewTableRowCount < dtNewTable.Rows.Count; iNewTableRowCount++)
{
filterCriterial = string.Empty;
foreach (string colName in columnNames.ToArray())
{
filterCriterial += colName.ToString() + "='" + dtNewTable.Rows[iNewTableRowCount][colName].ToString() + "' AND ";
}
filterCriterial = filterCriterial.TrimEnd((" AND ").ToCharArray());
DataRow[] dr = dtOriginalTable.Select(filterCriterial);
if (dr.Length > 0)
{
filterTable.Rows[filterTable.Rows.IndexOf(filterTable.Select(filterCriterial)[0])].Delete();
}
}
}
return filterTable;
}

How to sum columns in a dataTable?

How can I get a sum for all the columns in a datatable? Say I had the following table. How can I calculate the "total" row? It should be easy to add total row to a datatable.
Columns hits uniques sigups, etc...
Rows
1 12 1 23
2 1 0 5
3 6 2 9
total 19 3 37
Update
I ended up with this. It was the only thing I could get to work.
For Each col As DataColumn In TotalsTable.Columns
If col.DataType.Name = "DateTime" Then
count = count + 1
Continue For
End If
Dim colTotal As Double = 0
Dim value As Double
For Each row As DataRow In TotalsTable.Rows
If Double.TryParse(row(col), value) Then
colTotal += Double.Parse(row(col))
End If
Next
totalRow(count) = colTotal
count = count + 1
Next

There is also a way to do this without loops using the DataTable.Compute Method. The following example comes from that page. You can see that the code used is pretty simple.:
private void ComputeBySalesSalesID(DataSet dataSet)
{
// Presumes a DataTable named "Orders" that has a column named "Total."
DataTable table;
table = dataSet.Tables["Orders"];
// Declare an object variable.
object sumObject;
sumObject = table.Compute("Sum(Total)", "EmpID = 5");
}
I must add that if you do not need to filter the results, you can always pass an empty string:
sumObject = table.Compute("Sum(Total)", "")

Try this:
DataTable dt = new DataTable();
int sum = 0;
foreach (DataRow dr in dt.Rows)
{
foreach (DataColumn dc in dt.Columns)
{
sum += (int)dr[dc];
}
}

I doubt that this is what you want but your question is a little bit vague
Dim totalCount As Int32 = DataTable1.Columns.Count * DataTable1.Rows.Count
If all your columns are numeric-columns you might want this:
You could use DataTable.Compute to Sum all values in the column.
Dim totalCount As Double
For Each col As DataColumn In DataTable1.Columns
totalCount += Double.Parse(DataTable1.Compute(String.Format("SUM({0})", col.ColumnName), Nothing).ToString)
Next
After you've edited your question and added more informations, this should work:
Dim totalRow = DataTable1.NewRow
For Each col As DataColumn In DataTable1.Columns
totalRow(col.ColumnName) = Double.Parse(DataTable1.Compute("SUM(" & col.ColumnName & ")", Nothing).ToString)
Next
DataTable1.Rows.Add(totalRow)

You can loop through the DataColumn and DataRow collections in your DataTable:
// Sum rows.
foreach (DataRow row in dt.Rows) {
int rowTotal = 0;
foreach (DataColumn col in row.Table.Columns) {
Console.WriteLine(row[col]);
rowTotal += Int32.Parse(row[col].ToString());
}
Console.WriteLine("row total: {0}", rowTotal);
}
// Sum columns.
foreach (DataColumn col in dt.Columns) {
int colTotal = 0;
foreach (DataRow row in col.Table.Rows) {
Console.WriteLine(row[col]);
colTotal += Int32.Parse(row[col].ToString());
}
Console.WriteLine("column total: {0}", colTotal);
}
Beware: The code above does not do any sort of checking before casting an object to an int.
EDIT: add a DataRow displaying the column sums
Try this to create a new row to display your column sums:
DataRow totalsRow = dt.NewRow();
foreach (DataColumn col in dt.Columns) {
int colTotal = 0;
foreach (DataRow row in col.Table.Rows) {
colTotal += Int32.Parse(row[col].ToString());
}
totalsRow[col.ColumnName] = colTotal;
}
dt.Rows.Add(totalsRow);
This approach is fine if the data type of any of your DataTable's DataRows are non-numeric or if you want to inspect the value of each cell as you sum. Otherwise I believe #Tim's response using DataTable.Compute is a better.

It's a pity to use .NET and not use collections and lambda to save your time and code lines
This is an example of how this works:
Transform yourDataTable to Enumerable, filter it if you want , according a "FILTER_ROWS_FIELD" column, and if you want, group your data by a "A_GROUP_BY_FIELD".
Then get the count, the sum, or whatever you wish.
If you want a count and a sum without grouby don't group the data
var groupedData = from b in yourDataTable.AsEnumerable().Where(r=>r.Field<int>("FILTER_ROWS_FIELD").Equals(9999))
group b by b.Field<string>("A_GROUP_BY_FIELD") into g
select new
{
tag = g.Key,
count = g.Count(),
sum = g.Sum(c => c.Field<double>("rvMoney"))
};

for (int i=0;i<=dtB.Columns.Count-1;i++)
{
array(0, i) = dtB.Compute("SUM([" & dtB.Columns(i).ColumnName & "])", "")
}