Group on a datatable in a dataset 1

[stsuing]

I would like to return an array of rows from datatable. The Array returned should be the datatable rows grouped on certain fields.

Ex
Dataset table
1 3 124
2 3 444
3 3 444
4 3 555
5 4 123
6 4 123

Rows Returned
3 123
4 123
3 444
3 555


Something like
dim dRows() as datarow = _
dataset.tablename.Select(???)
or a dataview of the table.

Any help would be appreciated. I don't want to write an ugly loop

 

[jfrost10]

Couple of ways.

one was is to use the DataTable's Select method.

You pass it a filter (i.e. "id > 5&quot and a sorting description (i.e. "id DESC" or "id ASC&quot.

Another way is to dump your datatable into a DataView, and use it's sort method (a little simpler)
DataView1.Sort("id DESC&quot

Either way may be what you're looking for.

jack

 

[stsuing]

I'm not looking to sort the data. I'm looking to Group or select distinct rows on a table. The select method doesn't indicate that's possible.

Thanks anyway

 

[jfrost10]

ok. Question: why not just select distinct records when you initially fill the datatable from the database? do you need to have all those duplicate records hanging around?

Jack

 

[stsuing]

I need all the records because eventually they all make it to a file I'm writing. The structure of the file demand groupings based on those records. But you are correct. I could make another call to the database to get the groupings, but didn't want to.

It seems to me that If I had all the data I needed in a dataset then I should be able to come up with solution without making a call to the database.

 

[jfrost10]

k, give me a bit more bakground:
-what sort of file are you writing to?
-if all the records are for the file, what is the distinct list of records for? Selection? Display?
-what database are you using?

Jack

 

[stsuing]

I am writing a NSF 3.01 file and the detail lines contain information that will determine fields in the batch Header record, the number of batches, and the detail lines in those batches. I am using SQL 2000.

I think I've got a solution.

I will add a table to the dataset that has a column that contains the values that I wanted to group on. Then I will add a relation to the dataset, and then go through the table I added and getchildrows.

That's should work, and is better then crazy loops.

Thanks

 

[Zarcom]

You could also try this out. Basically you do a select on the data in the dataset and fill up your new datagrid with the new data. I tested this using the Northwinds database and it worked fine both grids I had on the page showed different data.

Private Sub Page_Load(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles MyBase.Load
SqlDataAdapter1.Fill(DataSet11)
DataGrid1.DataBind()
End Sub

Private Sub Button1_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles Button1.Click
Dim ds As New DataSet1()
ds.Merge(DataSet11.tablename.Select("sortexp&quot)

DataGrid2.DataSource = ds
DataGrid2.DataMember = "tablename"

DataGrid2.DataBind()

End Sub


HTH That'l do donkey, that'l do
Mark

 

[BG12424]

stsuing,

Do you have code that you can share for this? I am struggling with the same issue. I opened thread855-686103 which describes this same type of scenario.


Please read thread in this posting before continuing with this paragraph.>>
I'm not sure how I would be able to use the GetChildRows method in my case since I can't get unique email addresses when retrieving data. I do need to get a count of the unique addresses to be displayed in a label on the page. Can you shed any light on this? Can anyone? thanks so much.


regards,
Brian

http://www.gainesconsulting.com