LINQ performance Count vs Where and Count

Question

public class Group
{
   public string Name { get; set; }
}  

Test:

List _groups = new List();

for (i         


        

Answer 1

My guess:

.Where() uses special "WhereListIterator" to iterate over elements, Count() does not, as indicated by Wyatt Earp. The interesting thing is that the iterator is marked as "ngenable":

 [TargetedPatchingOptOut("Performance critical to inline this type of method across NGen image boundaries")]
 public WhereListIterator(List source, Func predicate)
 {
   this.source = source;
   this.predicate = predicate;
 }

This would probably mean that the "iterator" part runs as a "unmanaged code", while the Count() runs as a managed code. I don't know if that makes sense / how to prove it, but that's my 0.2cents.

Also, if you rewrite the Count() to take care of List carefully,

you can make it the same / even faster:

public static class TestExt{
   public static int CountFaster(this IEnumerable source, Func predicate) {
       if (source == null) throw new Exception();
       if (predicate == null) throw new Exception();

       if(source is List)
       {
                int finalCount=0;
                var list = (List)source;
                var count = list.Count;
                for(var j = 0; j < count; j++){
                    if(predicate(list[j])) 
                        finalCount++;
                }
                return finalCount;
       }


       return source.Count(predicate);
   }

}

On my tests; after I've started using CountFaster(), the one who is called LATER wins(because of cold-startup).