Develop Reference

Combining Where and OrderByDescending Using a Custom Method

I'm trying to use a custom method for ordering but I also want to use that same custom method to only return results that match a certain value. I realize that the code below works but I was hoping there was a way to combine both methods to hopefully speed up the process.
public IEnumerable<List<decimal>> GetBestList(List<List<decimal>> inputList)
{
var bestList = inputList.Where(x => x != null && CalculateAverage(x) > 0).
OrderByDescending(x => CalculateAverage(x)));
return bestList;
}
public decimal CalculateAverage(List<decimal> inputList)
{
return inputList.Average();
}

As far as I understand you want to prevent recalculation of average, so you can use Select to create a temporary tuple containing average and original list, for example like that:
public IEnumerable<List<decimal>> GetBestList(List<List<decimal>> inputList)
{
var bestList = inputList
.Where(x => x != null )
.Select(x => (x, Avg: CalculateAverage(x)))
.Where(x => x.Avg > 0)
.OrderByDescending(x => x.Avg)
.Select(x => x.x);
return bestList;
}

The way to avoid performing the potentially expensive computation multiple times is to project the sequence into a new value that includes the list and the computation. This is simpler and easier with query syntax than method syntax:
public IEnumerable<List<decimal>> GetBestList(List<List<decimal>> inputList)
{
var query = from list in inputList
where list != null
let average = CalculateAverage(list)
where average > 0
orderby average
select list;
}

Group by some columns depending on values in Entity Framework

I have the following simple statement in my Entity Framework code:
query = query
.Where(c => c.NotificationType == NotificationType.AppMessage)
.GroupBy(c => c.ConversationId)
.Select(d => d.OrderByDescending(p => p.DateCreated).FirstOrDefault());
It simply finds the latest Notification based on a group by with conversationId and select latest. Easy.
However, this is ONLY what I want if c.NotificationType == NotificationType.AppMessage. If the column is different than AppMessage (c.NotificationType <> NotificationType.AppMessage), I just want the column. What I truly Want to write is a magical statement such as:
query = query
.Where(c => (c.NotificationType <> NotificationType.AppMessage)
|| ((c.NotificationType == NotificationType.AppMessage)
.GroupBy(c => c.ConversationId)
.Select(d => d.OrderByDescending(p => p.DateCreated).FirstOrDefault()));
But this doesn't make sense because the GroupBy/Select is based on the first where statement.
How do I solve this?

The simplest way is to compose UNION ALL query using Concat at the end of your original query:
query = query
.Where(c => c.NotificationType == NotificationType.AppMessage)
.GroupBy(c => c.ConversationId)
.Select(d => d.OrderByDescending(p => p.DateCreated).FirstOrDefault())
.Concat(query.Where(c => c.NotificationType != NotificationType.AppMessage));

public class EntityClass
{
public int NotificationType { get; set; }
public int ConversationId { get; set; }
public DateTime Created { get; set; }
public static EntityClass GetLastNotification(int convId)
{
var list = new List<EntityClass>(); // Fill the values
list = list
.GroupBy(i => i.ConversationId) // Group by ConversationId.
.ToDictionary(i => i.Key, n => n.ToList()) // Create dictionary.
.Where(i => i.Key == convId) // Filter by ConversationId.
.SelectMany(i => i.Value) // Project multiple lists to ONLY one list.
.ToList(); // Create list.
// Now, you can filter it:
// 0 - NotificationType.AppMessage
// I didn't get what exactly you want to filter there, but this should give you an idea.
var lastNotification = list.OrderByDescending(i => i.Created).FirstOrDefault(i => i.NotificationType == 0);
return lastNotification;
}
}
you filter your list with "GroupBy" based on ConversationId. Next, create a dictionary from the result and make only one list (SelectMany). Then, you already have one list where should be only records with ConversationId you want.
Last part is for filtering this list - you wanted to last notification with certain NotificationType. Should be working :)

How to aggregate and SUM EntityFramework fields with multiple joins

I am able to produce a set of results that are desirable, but I have the need to group and sum of these fields and am struggling to understand how to approach this.
In my scenario, what would be the best way to get results that will:
Have a distinct [KeyCode] (right now I get many records, same KeyCode
but different occupation details)
SUM wage and projection fields (in same query)
Here is my LINQ code:
private IQueryable<MyAbstractCustomOccupationInfoClass> GetMyAbstractCustomOccupationInfoClass(string[] regionNumbers)
{
//Get a list of wage data
var wages = _db.ProjectionAndWages
.Join(
_db.HWOLInformation,
wages => wages.KeyCode,
hwol => hwol.KeyCode,
(wages, hwol) => new { wages, hwol }
)
.Where(o => regionNumbers.Contains(o.hwol.LocationID))
.Where(o => o.wages.areaID.Equals("48"))
.Where(o => regionNumbers.Contains(o.wages.RegionNumber.Substring(4))); //regions filter, remove first 4 characters (0000)
//Join OccupationInfo table to wage data, for "full" output results
var occupations = wages.Join(
_db.OccupationInfo,
o => o.wages.KeyCode,
p => p.KeyCode,
(p, o) => new MyAbstractCustomOccupationInfoClass
{
KeyCode = o.KeyCode,
KeyTitle = o.KeyTitle,
CareerField = o.CareerField,
AverageAnnualOpeningsGrowth = p.wages.AverageAnnualOpeningsGrowth,
AverageAnnualOpeningsReplacement = p.wages.AverageAnnualOpeningsReplacement,
AverageAnnualOpeningsTotal = p.wages.AverageAnnualOpeningsTotal,
});
//TO-DO: How to Aggregate and Sum "occupations" list here & make the [KeyCode] Distinct ?
return occupations;
}
I am unsure if I should perform the Grouping mechanism on the 2nd join? Or perform a .GroupJoin()? Or have a third query?

var occupations = _db.OccupationInfo.GroupJoin(
wages,
o => o.KeyCode,
p => p.wages.KeyCode,
(o, pg) => new MyAbstractCustomOccupationInfoClass {
KeyCode = o.KeyCode,
KeyTitle = o.KeyTitle,
CareerField = o.CareerField,
AverageAnnualOpeningsGrowth = pg.Sum(p => p.wages.AverageAnnualOpeningsGrowth),
AverageAnnualOpeningsReplacement = pg.Sum(p => p.wages.AverageAnnualOpeningsReplacement),
AverageAnnualOpeningsTotal = pg.Sum(p => p.wages.AverageAnnualOpeningsTotal),
});

LINQ: how to get an intersection of two sets of ints?

There must be a way to compare two sets of results while staying in LINQ. Here's my existing code that uses a HashSet to do the comparison after two separate queries:
public static void AssertDealershipsShareTransactionGatewayCredentialIds(long DealershipLocationId1,
long DealershipLocationId2)
{
using (var sqlDatabase = new SqlDatabaseConnection())
{
var DealershipCredentials1 =
sqlDatabase.Tables.DealershipLocationTransactionGateway
.Where(x => x.DealershipLocationId == DealershipLocationId1)
.Select(x => x.TransactionGatewayCredentialId);
var DealershipCredentials2 =
sqlDatabase.Tables.DealershipLocationTransactionGateway
.Where(x => x.DealershipLocationId == DealershipLocationId2)
.Select(x => x.TransactionGatewayCredentialId);
var doSetsOfCredentialsMatch = new HashSet<int>(DealershipCredentials1).SetEquals(DealershipCredentials2);
Assert.IsTrue(doSetsOfCredentialsMatch,
"The sets of TransactionGatewayCredentialIds belonging to each Dealership did not match");
}
}
Ideas? Thanks.

Easy answer (This will make 1, possibly 2 database calls, both of which only return a boolean):
if (list1.Except(list2).Any() || list2.Except(list1).Any())
{
... They did not match ...
}
Better answer (This will make 1 database call returning a boolean):
var DealershipCredentials1 =
sqlDatabase.Tables.DealershipLocationTransactionGateway
.Where(x => x.DealershipLocationId == DealershipLocationId1)
.Select(x => x.TransactionGatewayCredentialId);
var DealershipCredentials2 =
sqlDatabase.Tables.DealershipLocationTransactionGateway
.Where(x => x.DealershipLocationId == DealershipLocationId2)
.Select(x => x.TransactionGatewayCredentialId);
if (DealershipCredentials1.GroupJoin(DealershipCredential2,a=>a,b=>b,(a,b)=>!b.Any())
.Union(
DealershipCredentials2.GroupJoin(DealershipCredential1,a=>a,b=>b,(a,b)=>!b.Any())
).Any(a=>a))
{
... They did not match ...
}
The second method works by unioning a left outer join that returns a boolean indicating if any unmatching records were found with a right outer join that does the same. I haven't tested it, but in theory, it should return a simple boolean from the database.
Another approach, which is essentially the same as the first, but wrapped in a single LINQ, so it will always only make 1 database call:
if (list1.Except(list2).Union(list2.Except(list1)).Any())
{
}
And another approach:
var common=list1.Intersect(list2);
if (list1.Except(common).Union(list2.Except(common)).Any()) {}

Split query into multiple queries and then join the results

I have this function below that takes a list of id's and searches the DB for the matching persons.
public IQueryable<Person> GetPersons(List<int> list)
{
return db.Persons.Where(a => list.Contains(a.person_id));
}
The reason I need to split this into four queries is because the query can't take more than 2100 comma-separated values:
The incoming tabular data stream (TDS) remote procedure call (RPC) protocol stream is incorrect. Too many parameters were provided in this RPC request. The maximum is 2100.
How can I split the list into 4 pieces and make a query for each list. Then join the results into one list of persons?
Solved
I don't want to post it as an own answer and take cred away from #George Duckett's answer, just show the solution:
public IQueryable<Person> GetPersons(List<int> list)
{
var persons = Enumerable.Empty<Person>().AsQueryable<Person>();
var limit = 2000;
var result = list.Select((value, index) => new { Index = index, Value = value })
.GroupBy(x => x.Index / limit)
.Select(g => g.Select(x => x.Value).ToList())
.ToList();
foreach (var r in result)
{
var row = r;
persons = persons.Union(db.Persons.Where(a => row.Contains(a.person_id)));
}
return persons;
}

See this answer for splitting up your list: Divide a large IEnumerable into smaller IEnumerable of a fix amount of item
var result = list.Select((value, index) => new { Index = index, Value = value})
.GroupBy(x => x.Index / 5)
.Select(g => g.Select(x => x.Value).ToList())
.ToList();
Then do a foreach over the result (a list of lists), using the below to combine them.
See this answer for combining the results: How to combine Linq query results

I am not sure why you have a method like this. What exactly are you trying to do. Anyway you can do it with Skip and Take methods that are used for paging.
List<Person> peopleToReturn = new List<Person>();
int pageSize = 100;
var idPage = list.Skip(0).Take(pageSize).ToList();
int index = 1;
while (idPage.Count > 0)
{
peopleToReturn.AddRange(db.Persons.Where(a => idPage.Contains(a.person_id)).ToList());
idPage = list.Skip(index++ * pageSize).Take(pageSize).ToList();
}