如何计算两个单独的条件linq

Question

<person>
<name>Mehmet</name>
<date>25.07.1974</date>
<region>M</region>

Let's assume there is a XML file which contains and want to count
- persons who's birthdate is prior 2000 and
- person who's birthdate is prior 2000 and region M

I can get counts as below.

int x = List.Where( x=> x.Date < dateTime2000 ).Count();

int y = List.Where( x=> x.date < dateTime2000 && x.region == "M" ).Count();

Executions of aboves are fast.
But there are same comprasions and i feel it is not OK.
I do not calculate ToList() and ToArray() ,but i think above codes are more faster.
I am looking for faster alternative solution,if possible. Thanks for answers

Answer 1

You could use this:

var prior2000 = List.Where(x => x.Date < dateTime2000).ToList();
var x = prior2000.Count;
var y = prior2000.Count(x => x.region == "M");

This will only loop the results of the first query instead of all elements.

Answer 2

You can count how many items in x have region equal to "M"

 var itemsBefore2000 =  List.Where( x=> x.Date < dateTime2000 ).ToArray();
 int x = itemsBefore2000.Length;
 int y = itemsBefore2000.Where( x=> x.region == "M" ).ToArray().Length; 
         // or
         itemsBefore2000.Count( x=> x.region == "M" ); // this one is preferable

PS I wonder why so many people like to use Lists instead of arrays.

Answer 3

Well, you could at least reuse the first query for both counts:

var birthPrio2000 = List.Where( x=> x.Date < dateTime2000 ).ToList();
int countBirthPrio2000 = birthPrio2000.Count;
int countBirthPrio2000RegionM = birthPrio2000.Count(x => x.region == "M");

Another approach that uses a lookup (similar to a dictionary):

var birthPrio2000Regions = List.Where(x => x.Date < dateTime2000).ToLookup(x => x.Region);
int prio2000_TotalCount = birthPrio2000Regions.Sum(g => g.Count());
int prio2000_RegionM_Count = birthPrio2000Regions["M"].Count();

Answer 4

If you really want to ensure that the source is iterated only once you can use the .Aggregate() function and do the logic on yourself:

var desiredValues = persons.Aggregate(
    new { All = 0, InRegion = 0 }, 
    (sum, person) => new
    {
        All = person.Date < dateTime2000 ? sum.All + 1 : sum.All,
        InRegion = person.Region == "M" ? sum.InRegion + 1 : sum.InRegion
    });

var prio2000_TotalCount = desiredValues.All;
var prio2000_RegionM_Count = desiredValues.InRegion;