I'm having trouble getting my Linq statemnt to work when doing an outer join and a group by. Here's a SQL version of what I'm trying to accomplish:
select p.PRIMARY_KEY, min(p.EFFECTIVE_DATE), sum(IsNull(c.PAID_INDEMNITY, 0))
from PRMPOLCY p
left outer join CLMMAST c on p.PRIMARY_KEY = c.POLICY_NO
where p.UNDERWRITER_UID = 93
GROUP BY p.PRIMARY_KEY
Here's what I have in Linq (which doesn't work):
var result = from p in context.PRMPOLCies
join c in context.CLMMASTs on p.PRIMARY_KEY equals c.POLICY_NO into polClm
where (p.UNDERWRITER_UID == underwriter)
from grp in polClm.DefaultIfEmpty()
group grp by p.PRIMARY_KEY into g
select new PolicySummation()
{
PolicyNo = g.Key,
Incurred = g.Sum(grp => grp.PAID_INDEMNITY ),
EffDate = g.Min(grp => grp.PRMPOLCY.EFFECTIVE_DATE
};
Beating my head against the wall trying to figurwe this out!
Assuming you have a navigation property set up between PRMPOLCY and CLMMAST, you shouldn't need to specify the join explicitly. It's much easier to express most queries in linq without explicit joins, but rather treating your structures as a hierarchy. I don't know the specifics of your model property names, but I'd take a guess that something like this would work.
var result =
from p in context.PRMPOLCies
where (p.UNDERWRITER_UID == underwriter)
select new PolicySummation {
PolicyNo = p.PRIMARY_KEY,
Incurred = p.CLMASTs.Select(c => c.PAID_INDEMNITY).DefaultIfEmpty().Sum(),
EffDate = p.EFFECTIVE_DATE,
};
You need to include both your tables in the group clause like this:
group new { p, grp } by p.PRIMARY_KEY into g
Then in your Sum / Min
g.Sum(grp => grp.grp == null ? 0 : grp.grp.PAID_INDEMNITY )
g.Min(grp => grp.p.PRMPOLCY.EFFECTIVE_DATE)
Related
Following .net core EF core, Linq cannot be translated and will be evaluated locally. Can you please give me an advise?
var temp1= (from so in context.OrderShippingOrders
group so by so.OrderId into g
where g.Count(x=> x.IsSent == true ) == g.Count()
select new {
g.Key
}
);
query = (from o in context.Orders
join s in temp1
on o.Id equals s.Key
select o
);
The LINQ expression 'join AnonymousObject _o in {from Order o in value(Microsoft.EntityFrameworkCore.Query.Internal.EntityQueryable1[ECommerce.API.Models.Order]) where ([o].ShopId == __queryObj_ShopId_Value_0) join <>f__AnonymousType181 s in {from IGrouping2 g in {from OrderShippingOrder so in value(Microsoft.EntityFrameworkCore.Query.Internal.EntityQueryable1[ECommerce.API.Models.OrderShippingOrder]) orderby [so].OrderId asc, [so].OrderId asc select [so] => GroupBy([so].OrderId, [so])} where ({from OrderShippingOrder x in [g] where ([x].IsSent == True) select [x] => Count()} == {[g] => Count()}) select new <>f__AnonymousType181(Key = [g].Key)} on [o].Id equals [s].Key orderby EF.Property(?[o]?, "Id") asc select new AnonymousObject(new [] {Convert(EF.Property(?[o]?, "Id"), Object)}) => Skip(__p_1) => Take(__p_2) => Distinct()} on Property([o.OrderDetails], "OrderId") equals Convert([_o].GetValue(0), Nullable1)' could not be translated and will be evaluated locally.
If possible, upgrade to EF Core 2.1 (or 2.2) in order to get improved LINQ GroupBy translation.
Before version 2.1, in EF Core the GroupBy LINQ operator would always be evaluated in memory. We now support translating it to the SQL GROUP BY clause in most common cases.
There is nothing you can do in previous EF Core versions.
After upgrading, in order to get SQL transation, the GroupBy query must be modified to use intermediate projection and conditional Sum instead of conditional Count like this:
var temp1 = (from so in context.OrderShippingOrders
group new { SendCount = so.IsSent ? 1 : 0 } by so.OrderId into g
where g.Sum(x => x.SendCount) == g.Count()
select new
{
g.Key
}
);
(unfortunately the more natual group so and g.Sum(x => x.IsSent ? 1 : 0) does not translate, that's why we need the group new { SendCount = so.IsSent ? 1 : 0 } and g.Sum(x => x.SendCount))
P.S. In case you have collection navigation property from Order to OrderShippingOrder (something like public ICollection<OrderShippingOrder> Shipping { get; set; }), then you can avoid all these GroupBy complications and use simply:
var query = context.Orders
.Where(o => o.Shipping.Count(so => so.IsSent) == o.Shipping.Count());
I'm trying to convert my SQL statement to a Linq statement and I'm not sure how to add the second COUNT to it. This is my SQL statement
SELECT l.Campus_Name, Labs = COUNT(*), LabsWithSubnets = COUNT(s.Lab_Space_Id)
FROM vw_Lab_Space l
LEFT JOIN vw_Subnet s on l.Lab_Space_Id = s.Lab_Space_Id
GROUP BY l.Campus_Name
ORDER BY 1
and this is my LINQ statement so far:
from l in Vw_Lab_Space
from s in Vw_Subnet
.Where(s => s.Lab_Space_Id == l.Lab_Space_Id)
.DefaultIfEmpty() // <=- triggers the LEFT JOIN
group l by new { l.Campus_Name } into g
orderby g.Key.Campus_Name
select new {
Campus_Name = g.Key.Campus_Name,
Labs = g.Count()
}
So I have everything but the LabsWithSubnets part in there. I'm just not sure how to add that in as I can't just do an s.Lab_Space_id.Count() in the select statement.
If you need table structure and sample data please see Need help creating an OUTER JOIN to count spaces.
Using your query as a basis, you need the groups to include s so you can count when non-null (I also removed the unnecessary anonymous object around the grouping key):
from l in Vw_Lab_Space
from s in Vw_Subnet
.Where(s => s.Lab_Space_Id == l.Lab_Space_Id)
.DefaultIfEmpty() // <=- triggers the LEFT JOIN
group new { l, s } by l.Campus_Name into g
orderby g.Key
select new {
Campus_Name = g.Key,
Labs = g.Count(),
LabsWithSubnets = g.Count(ls => ls.s != null)
}
However, rather than translate the SQL, I would probably take advantage of LINQ's group join to handle the query slightly differently:
var ans = from l in Vw_Lab_Space
join s in Vw_Subnet on l.Lab_Space_Id equals s.Lab_Space_Id into sj
group new { l, sj } by ls.Campus_Name into lsjg
select new {
Campus_Name = lsjg.Key,
NumLabs = lsjg.Count(),
LabsWithSubnets = lsjg.Sum(lsj => lsj.sj.Count())
};
PS Even in your query, I would use join...from...DefaultIfEmpty rather than from...from...where but depending on your database engine, may not matter.
var abc1 = from dlist in db.DebtorTransactions.ToList()
join war in db.Warranties on dlist.ProductID equals war.Id
join ag in db.Agents on war.fldAgentID equals ag.pkfAgentID
join sr in db.SalesReps on war.fldSrId equals sr.pkfSrID
where dlist.TransTypeID == 1
select new
{
dlist.Amount,
dlist.TransTypeID,
name = ag.Name,
ag.pkfAgentID,
sr.pkfSrID,
salesnam = sr.Name
} into objabc
group objabc by new
{
objabc.TransTypeID,
objabc.name,
objabc.salesnam,
objabc.Amount
};
var amt1 = abc1.Sum(x => x.Key.Amount);
var abc2 = from dlist in db.DebtorTransactions.ToList()
join cjt in db.CarJackaTrackas on dlist.ProductID equals cjt.pkfCjtID
join ag in db.Agents on cjt.AgentID equals ag.pkfAgentID
join sr in db.SalesReps on cjt.SalesRepId equals sr.pkfSrID
where dlist.TransTypeID == 0
select new
{
dlist.Amount,
dlist.TransTypeID,
name = ag.Name,
ag.pkfAgentID,
sr.pkfSrID,
enter code here` salesnam = sr.Name
} into objabc
group objabc by new
{
objabc.TransTypeID,
objabc.name,
objabc.salesnam,
objabc.Amount
};
var amt2 = abc1.Sum(x => x.Key.Amount);
//var result1=
return View();
i am new to linq, this query is working but i need to get the sum of Amount where dlist.TransTypeID == 0 and where dlist.TransTypeID == 1 by just single query. may anybody help me? thanks in advance
Here's a trimmed down example of how you can do it. You can add the joins if they are necessary, but I'm not clear on why you need some of the extra join values.
var transTypeAmountSums = (from dlist in db.DebtorTransactions
group dlist by dlist.TransTypeId into g
where g.Key == 0 || g.Key == 1
select new
{
TransTypeId = g.Key,
AmountSum = g.Sum(d => d.Amount)
}).ToDictionary(k => k.TransTypeId, v => v.AmountSum);
int transTypeZeroSum = transTypeAmountSums[0];
int transTypeOneSum = transTypeAmountSums[1];
A couple of things to note:
I removed ToList(). Unless you want to bring ALL DebtorTransactions into memory then run a Linq operation on those results, you'll want to leave that out and let SQL take care of the aggregation (it's much better at it than C#).
I grouped by dlist.TransTypeId only. You can still group by more fields if you need that, but it was unclear in the example why they were needed so I just made a simplified example.
I am trying to create an outer join statement in LINQ and am not having much luck. I know that Performing an outer join requires two steps:
(1) Convert the join into a group join with into
(2) Use DefaultIfEmpty() on the group to generate the null value you expect if the joined result set is empty.
I have been using this code as an example:
var query = (from p in dc.GetTable<Person>()
join pa in dc.GetTable<PersonAddress>() on p.Id equals pa.PersonId into tempAddresses
from addresses in tempAddresses.DefaultIfEmpty()
select new { p.FirstName, p.LastName, addresses.State });
So I tried to do this:
var outerJoin =
from h in resultHours
join u in results on h.Key equals u.Key into outer
from dictionary in outer.DefaultIfEmpty()
select new {
The problem is that intellisense doesn't recognize anything in the select new {} statement. I've tried u. and h., even dictionary.
The problem I may be running into is that I'm trying to outer join two dictionaries. It doesn't seem to like that, although this is what I was told I need to do. I am obviously doing something wrong or not understanding something.
I need to join the dictionary results.Unit with the dictionary resultHours.Hours because my output is missing unitId fields under certain conditions. Doing an outer join is supposed to clear this up.
Here is the code for results:
var results =
(from v in VDimUnit
join vf in VFactEnergyAllocation on v.UnitKey equals vf.UnitKey
join vd in VDimGadsEvent on vf.GadsEventKey equals vd.GadsEventKey
join vt in VDimTime on vf.TimeKey equals vt.TimeKey
where typeCodes.Contains(vd.GadsEventTypeCode)
&& vt.YearNum >= (year - 3) && vt.YearNum <= year
group vf by new {v.PlantId, v.PhysicalUnitId, v.NetDependableCapacity, v.NetMaximumCapacity,
vt.MonthNum} into groupItem
select new {groupItem.Key.PlantId, groupItem.Key.PhysicalUnitId, groupItem.Key.NetMaximumCapacity,
groupItem.Key.MonthNum, PO_HRS = groupItem.Sum(
x=> (float)x.AllocatedEnergyMwh / groupItem.Key.NetDependableCapacity),
UO_HRS = groupItem.Sum(x=> (float)x.AllocatedEnergyMwh / groupItem.Key.NetDependableCapacity),
Unit = groupItem.Count(), groupItem.Key}).ToDictionary(x=> x.Key, x=> x);
Here is the code for resultHours:
var resultHours =
(from vt in VDimTime
join vf in VFactEnergyAllocation on vt.TimeKey equals vf.TimeKey
join vd in VDimGadsEvent on vf.GadsEventKey equals vd.GadsEventKey
join v in VDimUnit on vf.UnitKey equals v.UnitKey
group vt by new {v.PlantId, v.PhysicalUnitId, v.NetDependableCapacity, v.NetMaximumCapacity,
vt.MonthNum} into groupItem
select new {groupItem.Key.PlantId, groupItem.Key.PhysicalUnitId, groupItem.Key.NetMaximumCapacity,
Hours = groupItem.Count(), groupItem.Key}).ToDictionary(x=> x.Key.ToString(), x=> x.Hours);
This is presently how I have my output. It will change after I figure out how to do the outer join.
var finalResults =
(from r in results
orderby r.Key.MonthNum, r.Key.PlantId, r.Key.PhysicalUnitId
select new {Site = r.Key.PlantId, Unit = r.Key.PhysicalUnitId, r.Key.MonthNum, Numerator = r.Value.PO_HRS, Denominator =
resultHours[r.Key.ToString()], Weight = r.Key.NetMaximumCapacity, Data_Indicator = DATA_INDICATOR,
Budgeted = budgetedPlannedOutageHrs, Industry_Benchmark = INDUSTRY_BENCHMARK, Comments = comments,
Executive_Comments = executiveComments, Fleet_Exec_Comments = fleetExecComments});
I'm at a loss. The LINQ outer join examples I have found apaprently do not apply when joining dictionaries.
I've been looking through related LINQ questions here trying to figure this one out, but I'm having some trouble converting a SQL query of mine to the equivalent LINQ to Entities version.
select companies.CommpanyName,
job.Position,
count(offers.jobID) As Offered,
job.Openings,
job.Filled
from jobs
left outer join offers on jobs.ID = offers.JobID
join membership.dbo.individuals on jobs.UserID = individuals.ID
join membership.dbo.companies on individuals.CompanyID = companies.ID
where jobs.Hidden = 0
group by offers.JobID,
companies.CommpanyName,
job.Position,
job.Openings,
job.Filled
I've done left outer joins in LINQ before similar to this example but I'm not sure how to combine the count and group statements with this to get the desired result:
CompanyName Position Offered Openings Filled
1 Exmaple Co. Job X 0 2 0
2 Example Co. Job Y 4 6 3
3 Test Co. Job Z 1 1 1
The query is further complicated by the fact that it needs to utilize two separate data contexts. I apologize for the lack of example code, but I'm really not sure how to start this, my LINQ-fu is still weak.
Update:
This is the solution I arrived at with Craig's help, had to use LINQ to Objects because of the unfortunate multiple context setup, JobWithOfferCounts is not an entity object:
IEnumerable<Job> t = context1.JobSet.Include("Offers").Include("Contacts").Where(j => j.Hidden == false);
IEnumerable <JobWithOfferCounts> r = (from j in t
join i in context2.IndividualSet on j.UserID equals i.ID
join c in context2.CompanySet on i.CompanyID equals c.ID
select new JobWithOfferCounts()
{
JobObject = j,
CompanyID = Convert.ToInt32(c.ID),
CompanyName = c.HostName,
OfferCount = j.offers.Count(o => o.Rejected == false),
FilledCount = j.offers.Count(o => o.Accepted == true),
PendingCount = j.offers.Count(o => o.Accepted == false && o.Rejected == false)
});
return r;
I can't see why you have individuals in your query, or why you group by offers.JobID when it (unlike jobs.JobId) could be null, but here's a first stab:
var q = from c in Context.Companies
from i in c.Individuals
from j in i.Jobs
where j.Hidden == 0
select new
{
CompanyName = c.CompanyName,
Position = j.Position,
Offered = j.Offers.Count(),
Openings = j.Openings,
Filled = j.Filled
};
It's rarely correct to use join in LINQ to Entities or LINQ to SQL.