select distinct results using linq

Question

Situation

I have list of objects, and the linq query I am currently using is counting every item up twice. That happens because the data is setup that way. Every PO number will have two lines, and I am counting both. I only need to count one.

PO Number  Pallets
123          5
123          5
234          8
234          8
345          2
345          2

Linq Query

Pallets = (from o in this.OrdersModel
           where o.Pallets != null
           select double.Parse(o.Pallets))
           .Sum();

Assuming each object in the set is an OrderModel, when this query is run it will return 30 (all the rows added up)

Desired Outcome

The correct number of pallets would be 15, add up each PO pallets once.

Instead what I would like to happen is a way to select distinct po number, and then add up the pallets. It would be very easy to add this condition in straight SQL, but I have the slightest idea on how to do it in Linq.

Question

Is there a quick and easy way to do a select distinct po number using Linq?
If not, what would be the best way to accomplish something like this?

Do you mean distinct? There is a linq method called Distinct() although I use the lambda form so I'd struggle to advise on your query — Liath, Dec 18 '13 at 13:59
If you have two entries for each PO Number, how do you know which of the two to include in your sum? — Ian Nelson, Dec 18 '13 at 14:01
@IanNelson There are a few ways, but initially we do not have the data structure in place to need to worry about that yet. — zach, Dec 18 '13 at 14:49

Kamil Budziewski · Answer 1 · 2013-12-18T14:05:06.227

2

Try this:

Pallets = this.OrdersModel
                .Where(x=>x.Paletts!=null).Distinct()
                .Sum(x=>double.Parse(x.Pallets));

As here is said, you need to implement IEquatable interface, so you can define which items are duplicates (thanks to Raphaël for reminding)

edited Dec 18 '13 at 14:05

answered Dec 18 '13 at 14:00

Kamil Budziewski

22,699
14
85
105

1

Assuming `IEquatable` is implemented – Raphaël Althaus Dec 18 '13 at 14:02
@RaphaëlAlthaus yes of course, I forgot to mention that – Kamil Budziewski Dec 18 '13 at 14:04
@wudzik or use [Enumerable.Distinct Method (IEnumerable, IEqualityComparer)](http://msdn.microsoft.com/en-us/library/bb338049(v=vs.110).aspx) instead of implement `IEquatable` – Grundy Dec 18 '13 at 14:06
@Grundy yes, it's another valid option, but I prefer implementing IEquatable in such cases :) – Kamil Budziewski Dec 18 '13 at 14:07
@wudzik this one looks like it selects distinct palletts. I would need distinct po number – zach Dec 18 '13 at 15:25

score 2 · Accepted Answer · answered Dec 18 '13 at 14:06

2

Is there a quick and easy way to do a select distinct

You can use GroupBy:

Pallets = (from o in this.OrdersModel.GroupBy(o => o.PONumber).Select(g => g.First())
           where o.Pallets != null
           select double.Parse(o.Pallets))
           .Sum();

But you should probably try to fix your model so you don't get duplicate records in the first place!

answered Dec 18 '13 at 14:06

D Stanley

149,601
11
178
240

@D Stanley That is definately the best approach. We have not quite figured that step out yet. But we will need to eventually not populate every record. Thanks for the answer – zach Dec 18 '13 at 15:26

score 0 · Answer 3 · answered Dec 18 '13 at 14:03

0

Well, if you want a fast solution: Pallets = Pallets /2 But this is a quck fix , you should use distinct.

answered Dec 18 '13 at 14:03

PeterRing

1,767
12
20

select distinct results using linq

3 Answers3