Union of Two data Sets

tusharnarayan · October 30th, 2008, 06:00 AM

Hi all,

I want to take the union of two data sets. Both data set are the result of different query. So i cant't use the below syntax.

Select * from Sample where id = 2
Union
Select * from Sample where id = 3

Please provide me the alternate solution. One data set i am getting as the result of inflectional search and one is the result of Stored procedure.
Thanks in advance
Tushar

Old Pedant · October 30th, 2008, 03:24 PM

Well, you really would be better off doing the UNION in SQL, despite what you say. It would be *MUCH* more efficient. There's no reason you couldn't do a union based on those two results.

But anyway...

And you aren't really talking about union of two DataSet objects, I hope. I hope/assume you are talking about the union of two DataTable objects.

So it's really easy enough.

Just use AddRow method on the destination DataTable to add a row you get form the source DataTable.

You don't say what language you are using, so I'm not going to bother with a code example.

tusharnarayan · October 30th, 2008, 11:37 PM

Hi,

Thanks for your response.

I am using C#. As i said i want to take the union of two data sets in such a way that duplicate data should not be repeated.
Kindly reply me with code example.

Old Pedant · October 31st, 2008, 02:39 AM

One more time: NO, you are NOT taking the union of two data *SETS*. You are getting the union of two data *TABLES*. It's nonsense to talk about the union of data *SETS*.

Read what I wrote.

Code example tomorrow. Too late tonight.

tusharnarayan · October 31st, 2008, 03:24 AM

Ya, I forgot to mention for that. I am taking the union of two data tables only. Waiting for the code example.

Old Pedant · October 31st, 2008, 04:32 PM

Okay, *IF* the two DataTable objects are completely compatible--that is, same number and *TYPES* of columns, it's especially easy:

Code:

DataTable t1 = ...wherever you get this from...;
DataTable t2 = ...wherever you get this from...;

foreach(DataRow r2 in t2.Rows)
{
    t1.Rows.Add(r2);
}

If the schemas aren't the same, you'll have to create the new rows from the old ones on a field by field basis. Example:

Code:

DataTable t1 = ...wherever you get this from...;
DataTable t2 = ...wherever you get this from...;

foreach(DataRow r2 in t2.Rows)
{
    DataRow r1 = t1.NewRow();
    r1["name"] = r2["username"];
    r1["addr"] = r2["homeAddress"];
    r1["ssn"]  = r2["socialSecurityNumber"];
    rs["annualSalary"] = (double)(r2["intMonthlyWage"] * 12);
    t1.Rows.Add(r1);
}

Notice that you can alter the values and/or change types via cast as you do the copy if needed.

tusharnarayan · November 4th, 2008, 07:55 AM

Hi,

Thanks a lot for your response:). Now as per requirement i have to take the union of two list in such a way that there should not be any duplicates objects in the list... I have one property of objects so that i can compare the objects based on that property Say object.ID

Data type of List is Classobject.

Thanks in advance...
Tushar

Old Pedant · November 4th, 2008, 05:23 PM

Sorry, but now I am *SURE* you are making a big mistake.

Removing duplicates is *CLEARLY* a job for SQL, *NOT* for a DataTable or any other ADO.NET control.

Unless you have only a few records (less than one or two hundred) to be added from one table to the other, I can't possibly recommend doing this in .NET coding.

Yes, you *CAN* do this. You can filter the rows in DataTable1 based on the ID value from DataTable2. If you get no records from the filter, then you can add the record from 2 to 1. If you get one or more records, the ID is a duplicate, so it isn't added.

But this *will* be somewhat slow and clumsy compared to doing in SQL.