Linq "reporting" question with list properties - linq

I have an entity object that contains a list property. I'd like to expand the list values to the right. Being new to LINQ, I'm not sure how to do this. I could strongly type an object, but then I'd have to know the count/values at compile time and I'd like to make it more dynamic.
The output that I'm wanting is something like:
Name Demo1 Demo2 Demo3
Person Name1 TX TX
Person Name2 TX OK
Person Name3 TX TX OK
Main Class
public Main()
{
List<Event> events = new List<Event>();
events.Add(new Event()
{
EventDate = DateTime.Now,
EventLocation = new Location() { State = "TX" },
EventName = "Demo1"
});
events.Add(new Event()
{
EventDate = DateTime.Now,
EventLocation = events[0].EventLocation,
EventName = "Demo2"
});
events.Add(new Event()
{
EventDate = DateTime.Now,
EventLocation = new Location() { State = "OK" },
EventName = "Demo3"
});
List<Person> people = new List<Person>();
Person person1 = new Person();
person1.Name = "Person Name1";
person1.Events.Add(events[0]);
person1.Events.Add(events[1]);
Person person2 = new Person();
person2.Name = "Person Name2";
person2.Events.Add(events[0]);
person2.Events.Add(events[2]);
Person person3 = new Person();
person3.Name = "Person Name3";
person3.Events.Add(events[0]);
person3.Events.Add(events[1]);
person3.Events.Add(events[2]);
people.Add(person1);
people.Add(person2);
people.Add(person3);
}

It depends on whether you want to run the query in memory or in databse. In any case, you'll need to return a list with the "dynamic" part of the results, because you cannot dynamically generate members of anonymous types (and working with them would be difficult).
In memory (as in your example), you can write the following query:
// Find names of all events (for all people)
var allNames =
(from p in people
from e in p.Events
select e.EventName).Distinct();
// Find events for every person
var res =
from p in people
let known = p.Events.ToDictionary(e => e.EventName)
select new {
p.Name,
Events = allNames.Select(n =>
known.ContainsKey(n)?known[n].EventLocation:"N/A")
};
The first query gets names of all events (we use it later to find a value for all event for every person). The second query iterates over all people. It first creates dictionary with events (for fast lookup in memory) and then iterates over all event names and tries to find them in the dictionary (returning "N/A" if it is not found).

Related

decimal? with Left outer join gets null reference in LINQ

I am trying to do left outer join in LINQ for two vars but on selecting required coloumns, I get Object reference not set to an instance of an object error where I want Nullable decimal.
var FLS = (from ee in SumTillFYEnd
join es in SumTillFYStart on ee.Account equals es.Account into temp
from t in temp.DefaultIfEmpty()
select new
{
Account = ee.Account, // As of here it works
BeginDr = (t.DrStartCF == 0) ? (decimal?) null : t.DrStartCF // Here I get error Object reference not set to an instance of an object.
});
Some times SumTillFYEnd and some times SumTillFYStart becomes null. I want to join should work with default values, in case any one or both is null.
The problem is attempting to cast null to decimal?. You cannot ever directly cast null to another type, nullable or not. That will always cause a NullReferenceException. What you want instead is default. In other words, replace:
(decimal?)null
With
default(decimal?)
I solved this using a default class.
The reason I am seeing is that decimal can not be null so it either needs to set for a default value either 0 or decimal.MinValue
So, you require to have default class for SumTillFYStart like
var defaultSumTillFYStart = new SumTillFYStart { Account = string.Empty, DrStartCF =0};
With above in context, then in your piece of code replace
from t in temp.DefaultIfEmpty()
with this
from t in temp.DefaultIfEmpty(defaultSumTillFYStart)
I have a linqPad working written below but for different subset; I think it will help somebody:
void Main()
{
List<Debtor> debtors = new List<Debtor>();
List<SecurityHolding> holdings = new List<SecurityHolding>();
//Initialize Debtor
debtors.Add(new Debtor(){
AccountId = "J1",
OutstandingValue = 501.95M
});
debtors.Add(new Debtor(){
AccountId = "J2",
OutstandingValue = 75.68M
});
debtors.Add(new Debtor(){
AccountId = "J3",
OutstandingValue = 100.01M
});
//Initialize Security Holding
holdings.Add(new SecurityHolding(){
AccountId = "J2",
SecurityHoldingValue = 100M
});
holdings.Add(new SecurityHolding(){
AccountId = "J3",
SecurityHoldingValue = 200M
});
var defaultHolding = new SecurityHolding { AccountId= string.Empty, SecurityHoldingValue = 0};
var result = (from d in debtors
join p in holdings
on d.AccountId equals p.AccountId into temp
from t in temp.DefaultIfEmpty(defaultHolding)
select new
{
AccountId = d.AccountId,
OutstandingValue = d.OutstandingValue,
HoldingValue = (decimal?)t.SecurityHoldingValue
});
result.Dump();
}
// Define other methods and classes here
public class Debtor
{
public string AccountId {get;set;}
public decimal OutstandingValue {get;set;}
}
public class SecurityHolding
{
public string AccountId {get;set;}
public decimal SecurityHoldingValue {get;set;}
}
Here the output:

Entity Framework cycle of data

I have an Account object, which has many Transactions related to it.
In one method, I get all transactions for a particular account.
var transactionlines = (from p in Context.account_transaction
.Include("account_transaction_line")
// .Include("Account")
.Include("account.z_account_type")
.Include("account.institution")
.Include("third_party")
.Include("third_party.z_third_party_type")
.Include("z_account_transaction_type")
.Include("account_transaction_line.transaction_sub_category")
.Include("account_transaction_line.transaction_sub_category.transaction_category")
.Include("z_account_transaction_entry_type")
.Include("account_transaction_line.cost_centre")
where p.account_id == accountId
&& p.deleted == null
select p).ToList();
This is meant to return me a list of transactions, with their related objects. I then pass each object to a Translator, which translates them into data transfer objects, which are then passed back to my main application.
public TransactionDto TranslateTransaction(account_transaction source)
{
LogUserActivity("in TranslateTransaction");
var result = new TransactionDto
{
Id = source.id,
Version = source.version,
AccountId = source.account_id,
// Account = TranslateAccount(source.account, false),
ThirdPartyId = source.third_party_id,
ThirdParty = TranslateThirdParty(source.third_party),
Amount = source.transaction_amount,
EntryTypeId = source.account_transaction_entry_type_id,
EntryType = new ReferenceItemDto
{
Id = source.account_transaction_entry_type_id,
Description = source.z_account_transaction_entry_type.description,
Deleted = source.z_account_transaction_entry_type.deleted != null
},
Notes = source.notes,
TransactionDate = source.transaction_date,
TransactionTypeId = source.account_transaction_type_id,
TransactionType = new ReferenceItemDto
{
Id = source.z_account_transaction_type.id,
Description = source.z_account_transaction_type.description,
Deleted = source.z_account_transaction_type.deleted != null
}
};
... return my object
}
The problem is:
An account has Transactions, and a Transaction therefore belongs to an Account. It seems my translators are being called way too much, and reloading a lot of data because of this.
When I load my transaction object, it's 'account' property has a'transactions' propery, which has a list of all the transactions associated to that account. Each transaction then has an account property... and those account peroprties again, have a list of all the transactions... and on and on it goes.
Is there a way I can limit the loading to one level or something?
I have this set:
Context.Configuration.LazyLoadingEnabled = false;
I was hoping my 'Includes' would be all that is loaded... Don't load 'un-included' related data?
As requested, here is my TranslateAccount method:
public AccountDto TranslateAccount(account p, bool includeCardsInterestRateDataAndBalance)
{
LogUserActivity("in TranslateAccount");
if (p == null)
return null;
var result =
new AccountDto
{
Id = p.id,
Description = p.description,
PortfolioId = p.institution.account_portfolio_id,
AccountNumber = p.account_number,
Institution = TranslateInstitution(p.institution),
AccountType = new ReferenceItemDto
{
Id = p.account_type_id,
Description = p.z_account_type.description
},
AccountTypeId = p.account_type_id,
InstitutionId = p.institution_id,
MinimumBalance = p.min_balance,
OpeningBalance = p.opening_balance,
OpeningDate = p.opening_date
};
if (includeCardsInterestRateDataAndBalance)
{
// Add the assigned cards collection
foreach (var card in p.account_card)
{
result.Cards.Add(new AccountCardDto
{
Id = card.id,
AccountId = card.account_id,
Active = card.active,
CardHolderName = card.card_holder_name,
CardNumber = card.card_number,
ExpiryDate = card.expiry
});
}
// Populate the current interest rate
result.CurrentRate = GetCurrentInterestRate(result.Id);
// Add all rates to the account
foreach (var rate in p.account_rate)
{
result.Rates.Add(
new AccountRateDto
{
Id = rate.id,
Description = rate.description,
Deleted = rate.deleted != null,
AccountId = rate.account_id,
EndDate = rate.end_date,
Rate = rate.rate,
StartDate = rate.start_date
});
}
result.CurrentBalance = CurrentBalance(result.Id);
}
LogUserActivity("out TranslateAccount");
return result;
}
The entity framework context maintains a cache of data that has been pulled out of the database. Regardless of lazy loading being enabled/disabled, you can call Transaction.Account.Transactions[0].Account.Transactions[0]... as much as you want without loading anything else from the database.
The problem is not in the cyclical nature of entity framework objects - it is somewhere in the logic of your translation objects.

Find / Count Redundant Records in a List<T>

I am looking for a way to identify duplicate records...only I want / expect to see them.
So the records aren't duplicated completely but the unique fields I am unconcerned with at this point. I just want to see if they have made X# payments of the exact same amount, via the exact same card, to the exact same person. (Bogus example just to illustrate)
The collection is a List<> further whatever X# is the List<>.Count will be X#. In other words all the records in the list match (again just the fields I am concerned with) or I will reject it.
The best I can come up with is to take the first record get value of say PayAmount and LINQ the other two to see if they have the same PayAmount value. Repeat for all fields to be matched. This seems horribly inefficient but I am not smart enough to think of a better way.
So any thoughts, ideas, pointers would be greatly appreciated.
JB
Something like this should do it.
var duplicates = list.GroupBy(x => new { x.Amount, x.CardNumber, x.PersonName })
.Where(x => x.Count() > 1);
Working example:
class Program
{
static void Main(string[] args)
{
List<Entry> table = new List<Entry>();
var dup1 = new Entry
{
Name = "David",
CardNumber = 123456789,
PaymentAmount = 70.00M
};
var dup2 = new Entry
{
Name = "Daniel",
CardNumber = 987654321,
PaymentAmount = 45.00M
};
//3 duplicates
table.Add(dup1);
table.Add(dup1);
table.Add(dup1);
//2 duplicates
table.Add(dup2);
table.Add(dup2);
//Find duplicates query
var query = from p in table
group p by new { p.Name, p.CardNumber, p.PaymentAmount } into g
where g.Count() > 1
select new
{
name = g.Key.Name,
cardNumber = g.Key.CardNumber,
amount = g.Key.PaymentAmount,
count = g.Count()
};
foreach (var item in query)
{
Console.WriteLine("{0}, {1}, {2}, {3}", item.name, item.cardNumber, item.amount, item.count);
}
Console.ReadKey();
}
}
public class Entry
{
public string Name { get; set; }
public int CardNumber { get; set; }
public decimal PaymentAmount { get; set; }
}
The meat of which is this:
var query = from p in table
group p by new { p.Name, p.CardNumber, p.PaymentAmount } into g
where g.Count() > 1
select new
{
name = g.Key.Name,
cardNumber = g.Key.CardNumber,
amount = g.Key.PaymentAmount,
count = g.Count()
};
You're unique entries are based off of the 3 criteria of Name, Card Number, and Payment Amount so you group by them and then use .Count() to count how many of those unique values exist. where g.Count() > 1 filters the group to duplicates only.

Meaning of Payload-LINQ

I am new for LINQ, the above code is executing fine for me.
But I couldn't figure out why in the last sentence c.payload.Quantity is written instead of c.Quantity.
When I googled I came to know that the value Quantity is obtained from the payload that we get from the results in c, does that mean that the payload is the final output of filterQuery?
var orderData = new[]
{
new {OrderDate = new DateTime(2011,1,1,8,30,00), ItemID = "100", UnitPrice = 0.99, Quantity =5},
new {OrderDate = new DateTime(2011,1,1,8,31,00), ItemID = "200", UnitPrice = 3.99, Quantity =2},
new {OrderDate = new DateTime(2011,1,1,9,02,00), ItemID = "200", UnitPrice = 1.50, Quantity =1},
new {OrderDate = new DateTime(2011,1,1,9,07,00), ItemID = "100", UnitPrice = 4.10, Quantity =3}
};
var orders = orderData.ToPointStream(Application,ev=>PointEvent.CreateInsert(ev.OrderDate,ev),AdvanceTimeSettings.StrictlyIncreasingStartTime);
var filterQuery = from o in orders where o.Quantity>=2 select o;
(from o in filterQuery.ToIntervalEnumerable() where o.EventKind == EventKind.Insert
select new {o.StartTime,o.EndTime,o.Payload.Quantity}).Dump("Heading");
ToIntervalEnumerable transitions from the StreamInsight APIs and event model to a plain old .NET sequence. in the latter, in addition to the payload, you can access the system fields StartTime, EndTime, and EventKind. To avoid naming conflicts, the payload needs to be encapsulated (imagine the payload, which is defined by the user, had a field 'StartTime'!). this is the meaning of the Payload field.

EF single entity problem

I need to return a single instance of my viewmodel class from my repository in order to feed this into a strongly-typed view
In my repository, this works fine for a collection of viewmodel instances:
IEnumerable<PAWeb.Domain.Entities.Section> ISectionsRepository.GetSectionsByArea(int AreaId)
{
var _sections = from s in DataContext.Sections where s.AreaId == AreaId orderby s.Ordinal ascending select s;
return _sections.Select(x => new PAWeb.Domain.Entities.Section()
{
SectionId = x.SectionId,
Title = x.Title,
UrlTitle = x.UrlTitle,
NavTitle = x.NavTitle,
AreaId = x.AreaId,
Ordinal = x.Ordinal
}
);
}
But when I attempt to obtain a single entity, like this:
public PAWeb.Domain.Entities.Section GetSection(int SectionId)
{
var _section = from s in DataContext.Sections where s.SectionId == SectionId select s;
return _section.Select(x => new PAWeb.Domain.Entities.Section()
{
SectionId = x.SectionId,
Title = x.Title,
UrlTitle = x.UrlTitle,
NavTitle = x.NavTitle,
AreaId = x.AreaId,
Ordinal = x.Ordinal
}
);
}
I get
Error 1 Cannot implicitly convert type
'System.Linq.IQueryable<PAWeb.Domain.Entities.Section>' to
'PAWeb.Domain.Entities.Section'. An explicit conversion exists
(are you missing a cast?)"
This has got to be simple, but I'm new to c#, and I can't figure out the casting. I tried (PAWeb.Domain.Entities.Section) in various places, but no success. Can anyone help??
Your query is returning an IQueryable, which could have several items. For example, think of the difference between an Array or List of objects and a single object. It doesn't know how to convert the List to a single object, which one should it take? The first? The last?
You need to tell it specifically to only take one item.
e.g.
public PAWeb.Domain.Entities.Section GetSection(int SectionId)
{
var _section = from s in DataContext.Sections where s.SectionId == SectionId select s;
return _section.Select(x => new PAWeb.Domain.Entities.Section()
{
SectionId = x.SectionId,
Title = x.Title,
UrlTitle = x.UrlTitle,
NavTitle = x.NavTitle,
AreaId = x.AreaId,
Ordinal = x.Ordinal
}
).FirstOrDefault();
}
This will either return the first item, or null if there are no items that match your query. In your case that won't happen unless the table is empty since you don't have a where clause.

Resources