efficiently removing data without for loop

Question

I have following classes

public class Product
{
    public int Id { get; set; }
    public string Name { get; set; }
    public int CategoryId { get; set; }
    public int SectionId { get; set; }
    public string VendorName { get; set; }
}

public class ProductToRemove
{
    public int CategoryId { get; set; }
    public int SectionId { get; set; }
}

In the main I have list of these two classes as follows.

List<Product> Products = new List<Product>()
{
    new Product() { Id = 1, Name = "A", CategoryId = 11, SectionId = 6, VendorName = "ABC" },
    new Product() { Id = 2, Name = "B", CategoryId = 21, SectionId = 6, VendorName = "ABC" },
    new Product() { Id = 3, Name = "C", CategoryId = 13, SectionId = 8, VendorName = "ABC" },
    new Product() { Id = 4, Name = "D", CategoryId = 90, SectionId = 6, VendorName = "ABC" },
    new Product() { Id = 5, Name = "E", CategoryId = 25, SectionId = 9, VendorName = "ABC" },
};

    List<ProductToRemove> ProductsToRemove = new List<ProductToRemove>()
    {
        new ProductToRemove() {CategoryId = 11, SectionId = 6,  },
        new ProductToRemove() {CategoryId = 90, SectionId = 6,  }
    };

I want to remove anything from Products instance where CategoryId and SectionId matches what is in the ProductsToRemove collection. I know how to loop through Products collection and delete matching records but I am wondering if there is a way to do the same using Linq

Possible duplicate of [Remove item from list based on condition](http://stackoverflow.com/questions/3279145/remove-item-from-list-based-on-condition) — pookie, Jan 25 '17 at 20:19
LINQ would also use a loop -- it isn't magic. In fact, using LINQ here (with `GroupBy`/`Except` or what have you) would result in something that's less efficient than just looping yourself. If the list was sorted on `(CategoryId, SectionId)`, then a more efficient algorithm is possible -- but of course sorting consumes time itself. — Jeroen Mostert, Jan 25 '17 at 20:22

score 3 · Accepted Answer · answered Jan 25 '17 at 20:29

3

Products.RemoveAll(x => ProductsToRemove
    .Any(r => x.CategoryId == r.CategoryId && x.SectionId == r.SectionId));

answered Jan 25 '17 at 20:29

chadnt

1,095
9
24

Ehsan Sajjad · Answer 2 · 2017-01-25T20:30:35.263

1

You can use Where() in combination of Any() to achieve the required output:

var deletion = Products
   .Where(product=> ProductsToRemove
                   .Any(remove=> product.CategoryId == remove.CategoryId 
                                 && product.SectionId == remove.SectionId
                       )
          );

edited Jan 25 '17 at 20:30

answered Jan 25 '17 at 20:19

Ehsan Sajjad

61,834
16
105
160

the output should contain 3 rows with Product Ids 2, 3, 5. Is your code returning that? – jim crown Jan 25 '17 at 20:23
yes this is returning only two objects, those that will be removed but I already have that in ProductsToRemove – jim crown Jan 25 '17 at 20:29
just remove the `!` in start of it – Ehsan Sajjad Jan 25 '17 at 20:30

score 0 · Answer 3 · answered Jan 25 '17 at 20:26

List<Product> Products = new List<Product>()
{
   new Product() { Id = 1, Name = "A", CategoryId = 11, SectionId = 6, VendorName = "ABC" },
   new Product() { Id = 2, Name = "B", CategoryId = 21, SectionId = 6, VendorName = "ABC" },
   new Product() { Id = 3, Name = "C", CategoryId = 13, SectionId = 8, VendorName = "ABC" },
   new Product() { Id = 4, Name = "D", CategoryId = 90, SectionId = 6, VendorName = "ABC" },
   new Product() { Id = 5, Name = "E", CategoryId = 25, SectionId = 9,    VendorName = "ABC" },
};

List<Product> catsRemoved = Products.Where(x => x.CategoryId != 11 && x.CategoryId != 90).ToList();

I don't want to hard code these Ids in removing. It should be based on other List — jim crown, Jan 25 '17 at 20:28

score 0 · Answer 4 · answered Jan 25 '17 at 20:45

0

You could use the Linq except method: https://msdn.microsoft.com/en-us/library/bb300779(v=vs.110).aspx

answered Jan 25 '17 at 20:45

Theo

885
6
16

score 0 · Answer 5 · answered Feb 21 '19 at 19:08

For those who don't mind changing their data type.

If keys are not allowed to collide in your set than you can just use a normal Dictionary. Or implement Equals and use a HashSet (Note that HashSet is a set, not a list)

If keys may collide you can do this with a dictionary of lists.

Here is an example of what I mean.

Dictionary<string, List<Product>> Products = default;
List<Product> l = default.GetOrDefault(p.id1+"<|>"+p.id2); //LinQ extension on map
l.Clear();

Of course, you can always wrap this up in a class to disguise it as a List, probably with an actual list also in it to preserve order.

efficiently removing data without for loop

5 Answers5