Related
I have a financial application. I have ViewHistoricInstrumentValue which has rows like this
instrument1, date1, price, grossValue, netValue
instrument2, date1, price, grossValue, netValue
...
instrument1, date2, price, grossValue, netValue
...
My views are complicated but the db itself is small (4000 transactions). ViewHistoricInstrumentValue was executed in less than 1 second before I added the next CTE to the view. After that it takes 26s. ActualEvaluationPrice is the price for instrumentX at dateY. If this value is missing from HistoricPrice table then I find the previous price for instrumentX.
, UsedEvaluationPriceCte AS (
SELECT *
, isnull(ActualEvaluationPrice,
(select top 1 HistoricPrice.Price -- PreviousPrice
from HistoricPrice JOIN ValidDate
on HistoricPrice.DateId = ValidDate.Id
and HistoricPrice.InstrumentId = StartingCte.InstrumentId
and ValidDate.[Date] < StartingCte.DateValue
order by ValidDate.[Date]))
as UsedEvaluationPrice
FROM StartingCte
)
My problem is that the execution time increased needlessly. Right now the HistoricPrice table has no missing value so ActualEvaluationPrice is never null, so the previous price should be never determined.
ViewHistoricInstrumentValue returns 1815 rows. One other mystery is that the first query takes 26s, but the second only 2s.
SELECT * FROM [ViewHistoricInstrumentValue]
SELECT top(2000) * FROM [ViewHistoricInstrumentValue]
Appendix
The execution plan: https://www.dropbox.com/s/5st69uhjkpd3b5y/IsNull.sqlplan?dl=0
The same plan: https://www.brentozar.com/pastetheplan/?id=rk9bK1Wiv
The view:
ALTER VIEW [dbo].[ViewHistoricInstrumentValue] AS
WITH StartingCte AS (
SELECT
HistoricInstrumentValue.DateId
, ValidDate.Date as DateValue
, TransactionId
, TransactionId AS [Row]
, AccountId
, AccountName
, ViewTransaction.InstrumentId
, ViewTransaction.InstrumentName
, OpeningDate
, OpeningPrice
, Price AS ActualEvaluationPrice
, ClosingDate
, Amount
, isnull(ViewTransaction.FeeValue, 0) as FeeValue
, HistoricInstrumentValue.Id AS Id
FROM ViewBriefHistoricInstrumentValue as HistoricInstrumentValue
JOIN ValidDate on HistoricInstrumentValue.DateId = ValidDate.Id
JOIN ViewTransaction ON ViewTransaction.Id = HistoricInstrumentValue.TransactionId
left JOIN ViewHistoricPrice ON ViewHistoricPrice.DateId = HistoricInstrumentValue.DateId AND
ViewHistoricPrice.InstrumentId = ViewTransaction.InstrumentId
)
, UsedEvaluationPriceCte AS (
SELECT *
, isnull(ActualEvaluationPrice,
(select top 1 HistoricPrice.Price -- PreviousPrice
from HistoricPrice JOIN ValidDate
on HistoricPrice.DateId = ValidDate.Id
and HistoricPrice.InstrumentId = StartingCte.InstrumentId
and ValidDate.[Date] < StartingCte.DateValue
order by ValidDate.[Date]))
as UsedEvaluationPrice
FROM StartingCte
)
, GrossEvaluationValueCte AS (
SELECT *
, Amount * UsedEvaluationPrice AS GrossEvaluationValue
, (UsedEvaluationPrice - OpeningPrice) * Amount AS GrossCapitalGains
FROM UsedEvaluationPriceCte
)
, CapitalGainsTaxCte AS (
SELECT *
, dbo.MyMax(GrossCapitalGains * 0.15, 0) AS CapitalGainsTax
FROM GrossEvaluationValueCte
)
, IsOpenCte AS (
SELECT
DateId
, DateValue
, TransactionId
, [Row]
, AccountId
, AccountName
, InstrumentId
, InstrumentName
, OpeningDate
, OpeningPrice
, ActualEvaluationPrice
, UsedEvaluationPrice
, ClosingDate
, Amount
, GrossEvaluationValue
, GrossCapitalGains
, CapitalGainsTax
, FeeValue
, GrossEvaluationValue - CapitalGainsTax - FeeValue AS NetEvaluationValue
, GrossCapitalGains - CapitalGainsTax - FeeValue AS NetUnrealizedGains
, CASE WHEN ClosingDate IS NULL OR DateValue < ClosingDate
THEN CAST(1 AS BIT)
ELSE CAST(0 AS BIT)
END
AS IsOpen
, convert(NVARCHAR, DateValue, 20) + cast([Id] AS NVARCHAR(MAX)) AS Temp
, Id
FROM CapitalGainsTaxCte
)
Select * from IsOpenCte
I have no idea what your query is supposed to be doing. But this process:
ActualEvaluationPrice is the price for instrumentX at dateY. If this value is missing from HistoricPrice table then I find the previous price for instrumentX.
is handled easily with lag():
select vhiv.*
coalesce(vhiv.ActualEvaluationPrice,
lag(vhiv.ActualEvaluationPrice) over (partition by vhiv.InstrumentId order by DateValue)
) as UsedEvaluationPrice
from ViewHistoricInstrumentValue vhiv;
Note: If you need to filter out certain dates by joining to ValidDates, you can include the JOIN in the query. However, that is not part of the problem statement.
I need to be able to get rid of one of the workdate and one of the sr_name columns but I can't figure out how to.
I'm getting the query returned like this:
Query
I'm also getting this error message when entered into tableau:
The column 'sr_name' was specified multiple times for 'Custom SQL Query'.
Below is the code I have. If I remove a sr_name from either subquery there will be an error in the join clause.
select *
from
(
select s.sr_name, cast(punchdatetime as date) as workdate,
((datediff(second, min(case when p.InOut = 1 then punchdatetime end),
max(case when p.InOut = 0 then punchdatetime end))/3600) - .5) as
hoursworked
from PunchClock p join ServiceReps s on p.ServRepID = s.ServRepID
where punchyear >= 2019
group by s.sr_name, cast(punchdatetime as date)
) v
join
(
select sr_name, t.*,
calls = (select count(*) from CRM_Correspondence cr where
cast(cr.DateCreated as date) = workdate and StatusType like '%call%' and
cr.ServRepID = t.servrepid),
reaches = (select count(*) from CRM_Correspondence cr where
cast(cr.DateCreated as date) = workdate and (StatusType = 'call reached'
or StatusType like '%SCHEDULE%') and cr.ServRepID = t.servrepid),
books = (select count(*) from os_appointments o where cast(o.DateCreated
as date) = workdate and isnull(o.confirmedby, o.booked_by) =
t.servrepid),
attends = (select count(*) from os_appointments o where
cast(o.DateCreated as date) = workdate and isnull(o.confirmedby,
o.booked_by) = t.servrepid and o.appointmentStatus = 'attended')
from
(
select cast(cor.datecreated as date) workdate, cor.ServRepID
from CRM_Correspondence cor
where cor.datecreated > '2019-01-01'
group by cast(cor.datecreated as date), cor.servrepid
) t
join ServiceReps sr on t.ServRepID = sr.ServRepID
) u on v.sr_name = u.sr_name and v.workdate = u.workdate
I need the same results just without the duplicate column so I can enter the query into tableau.
This is challenging because you have so many subqueries here. This could be refactored to use a single query which is what I would do. But using the existing query you could do something along these lines.
I had to format this very differently so I could isolate each piece.
select v.sr_name
, v.workdate
, v.hoursworked
, u.ServRepID
, u.calls
, u.reaches
, u.books
, u.attends
from
(
select s.sr_name
, cast(punchdatetime as date) as workdate
, ((datediff(second, min(case when p.InOut = 1 then punchdatetime end), max(case when p.InOut = 0 then punchdatetime end))/3600) - .5) as hoursworked
from PunchClock p
join ServiceReps s on p.ServRepID = s.ServRepID
where punchyear >= 2019
group by s.sr_name
, cast(punchdatetime as date)
) v
join
(
select sr_name
, t.*
, calls =
(
select count(*)
from CRM_Correspondence cr
where cast(cr.DateCreated as date) = workdate
and StatusType like '%call%'
and cr.ServRepID = t.servrepid
)
, reaches =
(
select count(*)
from CRM_Correspondence cr
where cast(cr.DateCreated as date) = workdate
and (StatusType = 'call reached' or StatusType like '%SCHEDULE%')
and cr.ServRepID = t.servrepid
)
, books =
(
select count(*)
from os_appointments o
where cast(o.DateCreated as date) = workdate and isnull(o.confirmedby, o.booked_by) = t.servrepid
)
, attends =
(
select count(*)
from os_appointments o
where cast(o.DateCreated as date) = workdate
and isnull(o.confirmedby, o.booked_by) = t.servrepid
and o.appointmentStatus = 'attended'
)
from
(
select cast(cor.datecreated as date) workdate
, cor.ServRepID
from CRM_Correspondence cor
where cor.datecreated > '2019-01-01'
group by cast(cor.datecreated as date)
, cor.servrepid
) t
join ServiceReps sr on t.ServRepID = sr.ServRepID
) u on v.sr_name = u.sr_name
and v.workdate = u.workdate
The SQL below is written to return 'open orders'. This was written in a way I could understand but now I would like to try and optimize this and reduce the amount of code.
The SQL below gives me the desired outcome I'm looking for however, I would like to shorten the query without using WITH AS. Any suggestions using UNION or some other nesting method?
WITH
product AS --filter by dept
(SELECT item
, dept
FROM item_master
WHERE dept in ('353')
),
open_orders AS --view of orders in Status A with ordered units > received units
(SELECT ol.order_no
, ol.item
, ol.location
, oh.po_type
, oh.order_type
, oh.not_before_date
, oh.not_after_date
, oh.otb_eow_date
, SUM(ol.qty_ordered) AS qty_ordered
, SUM(NVL(ol.qty_received,0)) AS qty_received
FROM ordhead oh
, ordloc ol
WHERE oh.order_no = ol.order_no
AND oh.status = 'A'
AND ol.qty_ordered > NVL(ol.qty_received,0)
-- AND ol.order_no in ('18701212') --optional filter for specific PO's
GROUP BY ol.order_no
, ol.item
, ol.location
, oh.po_type
, oh.order_type
, oh.not_before_date
, oh.not_after_date
, oh.otb_eow_date
),
allocations AS --view of all allocations
(SELECT ah.alloc_no
, ah.order_no
, ah.item
, ad.to_loc
, NVL(ad.qty_allocated,0) AS qty_allocated
, NVL(ad.qty_received,0) AS qty_received
FROM alloc_header ah
, alloc_detail ad
WHERE ah.alloc_no = ad.alloc_no
)
SELECT p.dept --main query on above views
, oo.order_no
, oo.po_type
, oo.order_type
, oo.not_before_date
, oo.not_after_date
, oo.otb_eow_date
, oo.item
, CASE WHEN oo.po_type = 0 THEN oo.location ELSE aa.to_loc END AS loc
, SUM(oo.qty_ordered) AS order_qty
, CASE WHEN SUM(NVL(aa.qty_allocated,0)) - SUM(NVL(aa.qty_received,0)) = 0
THEN SUM(oo.qty_ordered) - SUM(NVL(oo.qty_received,0))
ELSE SUM(NVL(aa.qty_allocated,0)) - SUM(NVL(aa.qty_received,0))
END AS open_qty
FROM open_orders oo
, allocations aa
, product p
WHERE oo.order_no = aa.order_no(+)
AND oo.item = aa.item(+)
AND oo.item = p.item
AND (oo.qty_ordered - oo.qty_received) >0
GROUP BY p.dept
, oo.order_no
, oo.po_type
, oo.order_type
, oo.not_before_date
, oo.not_after_date
, oo.otb_eow_date
, oo.item
, CASE WHEN oo.po_type = 0 THEN oo.location ELSE aa.to_loc END
;
CTE's (Common Table Expressions) are just a way of organizing a query by sticking bits of code (that define a "derived" table) at the top that can be reused in the main statement. As such, where product, open_orders, and allocations are mentioned in the FROM clause, you can just swap those words out with the code that defines them:
SELECT p.dept --main query on above views
,
oo.order_no,
oo.po_type,
oo.order_type,
oo.not_before_date,
oo.not_after_date,
oo.otb_eow_date,
oo.item,
CASE
WHEN oo.po_type = 0
THEN oo.location
ELSE aa.to_loc
END AS loc,
SUM(oo.qty_ordered) AS order_qty,
CASE
WHEN SUM(NVL(aa.qty_allocated, 0)) - SUM(NVL(aa.qty_received, 0)) = 0
THEN SUM(oo.qty_ordered) - SUM(NVL(oo.qty_received, 0))
ELSE SUM(NVL(aa.qty_allocated, 0)) - SUM(NVL(aa.qty_received, 0))
END AS open_qty
FROM (
SELECT ol.order_no,
ol.item,
ol.location,
oh.po_type,
oh.order_type,
oh.not_before_date,
oh.not_after_date,
oh.otb_eow_date,
SUM(ol.qty_ordered) AS qty_ordered,
SUM(NVL(ol.qty_received, 0)) AS qty_received
FROM ordhead oh,
ordloc ol
WHERE oh.order_no = ol.order_no
AND oh.STATUS = 'A'
AND ol.qty_ordered > NVL(ol.qty_received, 0)
-- AND ol.order_no in ('18701212') --optional filter for specific PO's
GROUP BY ol.order_no,
ol.item,
ol.location,
oh.po_type,
oh.order_type,
oh.not_before_date,
oh.not_after_date,
oh.otb_eow_date
) oo,
(
SELECT ah.alloc_no,
ah.order_no,
ah.item,
ad.to_loc,
NVL(ad.qty_allocated, 0) AS qty_allocated,
NVL(ad.qty_received, 0) AS qty_received
FROM alloc_header ah,
alloc_detail ad
WHERE ah.alloc_no = ad.alloc_no
) aa,
(
SELECT item,
dept
FROM item_master
WHERE dept IN ('353')
) p
WHERE oo.order_no = aa.order_no(+)
AND oo.item = aa.item(+)
AND oo.item = p.item
AND (oo.qty_ordered - oo.qty_received) > 0
GROUP BY p.dept,
oo.order_no,
oo.po_type,
oo.order_type,
oo.not_before_date,
oo.not_after_date,
oo.otb_eow_date,
oo.item,
CASE
WHEN oo.po_type = 0
THEN oo.location
ELSE aa.to_loc
END;
This is obviously not shortened (but by a few characters), but I get the sense that "shortening" isn't your requirement. You are trying to get this query to work in a product that doesn't support CTEs.
I am working on a SQL Query to group the results of a View by Id to have only one Row per Id
with a maximum of three pivoted results and keeping some columns static : TestCaseId, TestName, Test Case Num, Owner
Actually this is the Query i Created to get the Desired output but is now working as expected because the MAX is always retriving the max value so i am getting only one row but the pivoted values are repeated to the right.
SELECT DISTINCT TBL1.[TestName], TBL1.[Test Case Num], TBL1.[Owner], MAX(TBL1.[Browser]) as 'Column1', MAX(TBL1.[Run Date]) as 'Column2', MAX(TBL1.[Status]) as 'Column3', MAX(TBL1.[Duration]) as 'Column4', MAX(TBL1.[ErrorMsg]) as 'Column5', MAX(TBL2.[Browser]) as 'Column6', MAX(TBL2.[Run Date]) as 'Column7', MAX(TBL2.[Status]) as 'Column8', MAX(TBL2.[Duration]) as 'Column9', MAX(TBL2.[ErrorMsg]) as 'Column10', MAX(TBL3.[Browser]) as 'Column11' , MAX(TBL3.[Run Date]) as 'Column12', MAX(TBL3.[Status]) as 'Column13', MAX(TBL3.[Duration]) as 'Column14', MAX(TBL3.[ErrorMsg]) as 'Column15'
FROM (SELECT DISTINCT T1.[TestCaseId], T1.[TestName], T1.[Test Case Num], T1.[Owner], T1.[Browser], T1.[Run Date], T1.[Status], T1.[Duration], T1.[ErrorMsg]
FROM [TestRunner].[dbo].RunsRawResults T1) TBL1
cross apply (SELECT DISTINCT T2.[TestCaseId], T2.[Browser], T2.[Run Date], T2.[Status], T2.[Duration], T2.[ErrorMsg]
FROM [TestRunner].[dbo].RunsRawResults T2
WHERE T2.[TestCaseId] = TBL1.[TestCaseId] AND T2.[Run Date] TBL1.[Run Date]) TBL2
cross apply (SELECT DISTINCT T3.[TestCaseId], T3.[Browser], T3.[Run Date], T3.[Status], T3.[Duration], T3.[ErrorMsg]
FROM [TestRunner].[dbo].RunsRawResults T3
WHERE T3.[TestCaseId] = TBL2.[TestCaseId] AND T3.[Run Date] TBL2.[Run Date] AND T3.[Run Date] TBL1.[Run Date]) TBL3
GROUP BY TBL1.[TestCaseId], TBL1.[TestName], TBL1.[Test Case Num], TBL1.[Owner]
Input -
Raw Data (Comes from the RunRawResults View)
Desired and Pivoted Output
Using a common table expression (cte) and row_number() we can simplify the identification and order of multiple run dates. This also lets us skip using distinct and group by.
Switching to outer apply lets us include results where there are less than 3 runs per TestCaseId.
;with cte as (
select *
, rn = row_number() over (
partition by TestCaseId
order by [Run Date]
)
from TestRunner.dbo.RunsRawResults
)
select
tbl1.TestName
, tbl1.[Test Case Num]
, tbl1.Owner
, tbl1.Browser
, [Run Date]_tbl1 = tbl1.[Run Date]
, Status_tbl1 = tbl1.Status
, Duration_tbl1 = tbl1.Duration
, ErrorMsg_tbl1 = tbl1.ErrorMsg
, Browser_tbl2 = tbl2.Browser
, [Run Date]_tbl2 = tbl2.[Run Date]
, Status_tbl2 = tbl2.Status
, Duration_tbl2 = tbl2.Duration
, ErrorMsg_tbl2 = tbl2.ErrorMsg
, Browser_tbl3 = tbl3.Browser
, [Run Date]_tbl3 = tbl3.[Run Date]
, Status_tbl3 = tbl3.Status
, Duration_tbl3 = tbl3.Duration
, ErrorMsg_tbl3 = tbl3.ErrorMsg
from cte as tbl1
outer apply (
select
i.[Run Date]
, i.Status
, i.Duration
, i.ErrorMsg
from cte as i
where i.TestCaseId = tbl1.TestCaseId
and i.rn = 2
) as tbl2
outer apply (
select
i.[Run Date]
, i.Status
, i.Duration
, i.ErrorMsg
from cte as i
where i.TestCaseId = tbl1.TestCaseId
and i.rn = 3
) as tbl3
where tbl1.rn = 1
This is a follow up from this post. What I am trying to do now is sum the total payments made for the last 6 months. For example, we have this loan
as you can see they made 3 payments in the month of April, what I need to do is sum those to get the net amount. Currently my query just finds one of them and takes that one but that is not correct. What I tried to do is this:
payments as
(
SELECT ROW_NUMBER() OVER(Partition By Account ORDER BY CONVERT(datetime,DateRec) DESC) AS [RowNumber], Total as [Total], Account, SourceTyp
FROM mars.dbo.vw_PaymentHistory
WHERE SourceTyp = 'RegPmt'
GROUP BY Total, Account, DateRec, SourceTyp
)
,
get_payment1 as
(
SELECT * FROM payments
where RowNumber = 1 AND SourceTyp = 'RegPmt'
)
,
get_payment2 as
(
SELECT * FROM payments
where RowNumber = 2 AND SourceTyp = 'RegPmt'
),
get_payment3 as
(
SELECT * FROM payments
where RowNumber = 3 AND SourceTyp = 'RegPmt'
),
get_payment4 as
(
SELECT * FROM payments
where RowNumber = 4 AND SourceTyp = 'RegPmt'
),
get_payment5 as
(
SELECT * FROM payments
where RowNumber = 5 AND SourceTyp = 'RegPmt'
),
get_payment6 as
(
SELECT * FROM payments
where RowNumber = 6 AND SourceTyp = 'RegPmt'
)
But for the loan above I only get the negative value and what I need is the sum of all the payments made for the month which is specified by DateRec. I am unsure what to do at this point to achieve the result.
Here is my entire query I have so far:
Declare #snapshotdate date = '5/12/2020',
#monthStart date = '5/1/2020';
WITH Active_Loans as (
SELECT
la.Account,
la.LoanStatus,
la.PrinBal,
isnull(b.Amount, 0) [DUPB],
la.PrinBal + isnull(b.Amount, 0) [TotalUPB],
l.NoteOwner,
pt.[Partition],
l.paidoffdate,
la.[First Name],
la.[Last Name],
la.PmtPI,
la.PmtImpound,
la.NextDueDate,
la.MaturityDate,
la.NoteOwner as [Note Owner]
FROM MARS_DW..vw_Loans_Archive la
LEFT JOIN MARS_DW..vw_DUPBByLoan b on b.Account = la.Account
AND b.ArchiveDate = la.ArchiveDate
LEFT JOIN MARS..vw_Loans l on l.Account = la.Account
LEFT JOIN Portfolio_Analytics..partition_table pt on pt.Noteowner = l.NoteOwner
WHERE la.ArchiveDate = #snapshotdate
AND la.isActive = 1
--AND la.PaidOffDate is null
--AND la.LoanStatus NOT LIKE 'BK Payment Plan'
--AND la.LoanStatus NOT LIKE 'Prelim'
--AND la.LoanStatus NOT like 'trailing claims'
--AND la.Account NOT IN (
-- SELECT account
-- FROM MARS..vw_Loans
-- WHERE servicexferdate <=
-- DATEADD(dd, - 1, DATEADD(mm, DATEDIFF(mm, 0, #monthStart) + 1, 0))
-- AND PaidOffDate BETWEEN #monthStart AND DATEADD(dd, - 1, DATEADD(mm, DATEDIFF(mm, 0, #monthStart) + 1, 0))
-- )
UNION
(
SELECT l.account
,la.LoanStatus
,la.PrinBal
,isnull(b.Amount, 0) [DUPB]
,la.PrinBal + isnull(b.Amount, 0) [TotalUPB]
,l.NoteOwner
,pt.[Partition]
,l.PaidOffDate
,la.[First Name]
,la.[Last Name]
,la.PmtPI
,la.PmtImpound
,la.NextDueDate
,la.MaturityDate
,la.NoteOwner as [Note Owner]
FROM MARS..vw_Loans l
LEFT JOIN MARS_DW..vw_Loans_Archive la on la.Account = l.Account
LEFT JOIN MARS_DW..vw_DUPBByLoan b on b.Account = la.Account
LEFT JOIN Portfolio_Analytics..partition_table pt on pt.Noteowner = l.NoteOwner
AND b.ArchiveDate = la.ArchiveDate
WHERE l.servicexferdate < #snapshotdate
AND l.PaidOffDate > #snapshotdate
AND la.ArchiveDate = #snapshotdate
AND la.LoanStatus NOT like 'trailing claims'
AND la.LoanStatus NOT like 'Inactive - REO/FC'
AND pt.[Partition] IS NOT NULL
)
)
,
payments as
(
SELECT ROW_NUMBER() OVER(Partition By Account ORDER BY CONVERT(datetime,DateRec) DESC) AS [RowNumber], Total as [Total], Account, SourceTyp
FROM mars.dbo.vw_PaymentHistory
WHERE SourceTyp = 'RegPmt'
GROUP BY Total, Account, DateRec, SourceTyp
)
,
get_payment1 as
(
SELECT * FROM payments
where RowNumber = 1 AND SourceTyp = 'RegPmt'
)
,
get_payment2 as
(
SELECT * FROM payments
where RowNumber = 2 AND SourceTyp = 'RegPmt'
),
get_payment3 as
(
SELECT * FROM payments
where RowNumber = 3 AND SourceTyp = 'RegPmt'
),
get_payment4 as
(
SELECT * FROM payments
where RowNumber = 4 AND SourceTyp = 'RegPmt'
),
get_payment5 as
(
SELECT * FROM payments
where RowNumber = 5 AND SourceTyp = 'RegPmt'
),
get_payment6 as
(
SELECT * FROM payments
where RowNumber = 6 AND SourceTyp = 'RegPmt'
)
SELECT
rptpop.Account
, rptpop.LoanStatus
, rptpop.[First Name]
, rptpop.[Last Name]
, '$' + CONVERT (VARCHAR (12), rptpop.PmtPI+rptpop.PmtImpound, 1) as PITI
,'$' + CONVERT (VARCHAR (12), rptpop.TotalUPB, 1) as [Total UPB]
, CONVERT(VARCHAR(10),rptpop.NextDueDate,101) as [Next Due Date]
, CONVERT(VARCHAR(10),rptpop.MaturityDate,101) as [Maturity Date]
, rptpop.[Note Owner]
, '$' + CONVERT (VARCHAR (12),c1.Total , 1) as [c1]
, '$' + CONVERT (VARCHAR (12),c2.Total , 1) as [c2]
, '$' + CONVERT (VARCHAR (12),c3.Total , 1) as [c3]
, '$' + CONVERT (VARCHAR (12),c4.Total , 1) as [c4]
, '$' + CONVERT (VARCHAR (12),c5.Total , 1) as [c5]
, '$' + CONVERT (VARCHAR (12),c6.Total , 1) as [c6]
FROM Active_Loans as rptpop
LEFT JOIN get_payment1 as c1 on c1.Account = rptpop.Account
LEFT JOIN get_payment2 as c2 on c2.Account = rptpop.Account
LEFT JOIN get_payment3 as c3 on c3.Account = rptpop.Account
LEFT JOIN get_payment4 as c4 on c4.Account = rptpop.Account
LEFT JOIN get_payment5 as c5 on c5.Account = rptpop.Account
LEFT JOIN get_payment6 as c6 on c6.Account = rptpop.Account
WHERE
rptpop.Partition = 'GAEA'
AND rptpop.[Last Name] NOT LIKE '%CRE%'
If something was not clear with what I was asking please let me know. Again I just want to create a way to find all payments made in 6 months for each loan, I would assume it just needs to be a slight modification that was done in the previous post I made.
Edit:
I have tried something like this as well:
payments as
(
SELECT ROW_NUMBER() OVER(Partition By Account ORDER BY CONVERT(datetime,DateRec) DESC) AS [RowNumber], Account
, SUM(Total) OVER(Partition By Account ORDER BY CONVERT(datetime,DateRec) DESC) AS Total
FROM mars.dbo.vw_PaymentHistory
WHERE SourceTyp = 'RegPmt'
)
,
But I get this error
Msg 102, Level 15, State 1, Line 76
Incorrect syntax near 'order'.
Msg 102, Level 15, State 1, Line 87
Incorrect syntax near ','.
Msg 102, Level 15, State 1, Line 93
Incorrect syntax near ','.
Msg 102, Level 15, State 1, Line 99
Incorrect syntax near ','.
Msg 102, Level 15, State 1, Line 105
Incorrect syntax near ','.
Msg 102, Level 15, State 1, Line 111
Incorrect syntax near ','.
Edit
To make this really clear this is the output that I should get:
This is too long for a comment
SELECT ROW_NUMBER() OVER(Partition By Account ORDER BY CONVERT(datetime,DateRec) DESC) AS [RowNumber], Total as [Total], Account, SourceTyp
FROM mars.dbo.vw_PaymentHistory
WHERE SourceTyp = 'RegPmt'
GROUP BY Total, Account, DateRec, SourceTyp
when you have two payments for the same Account, Total, DateRec and SourceTyp, you take it as one payment? That does not seem right. Also, if you have ties on DateRec for some Account, the ROW_NUMBER() assignment for the ties is not deterministic. And that may be a problem, because you evaluate the CTE six times and you may end up with duplicate or missing values in the result. You should add columns to the ORDER BY to make it deterministic.
For the question, maybe another CTE?
get_payment6sum as
(
SELECT sum(Total) as Last6PaymentsSum, Account
FROM payments
WHERE RowNumber <= 6 AND SourceTyp = 'RegPmt'
GROUP BY Account
)
You should start with summarizing each months payment and assign them row_numbers or filter on last 6 months depending on requirement. You can get last 6 payments using the pseudo code below.
WITH payments as
(
SELECT
ROW_NUMBER() OVER(Partition By Account ORDER BY CONVERT(VARCHAR(6),DateRec,112) DESC) as rn,
'C' + Convert(char(4),ROW_NUMBER() OVER(Partition By Account ORDER BY CONVERT(VARCHAR(6),DateRec,112) DESC)) AS [pmnt_Number],
Account,
CONVERT(VARCHAR(6),DateRec,112) as pmnt_month,
SUM(Total) AS Total_payment
FROM tbl_PaymentHistory
WHERE SourceTyp = 'RegPmt'
GROUP BY Account, CONVERT(VARCHAR(6),DateRec,112)
)
Select * from payments WHERE rn <= 6
Once you have your last 6 months payments for your accounts with other attributes then you can do a pivot to get an output like you mentioned or you can also use the left join approach but you can simply use payments CTE instead of creating 6 CTEs.
WITH PIVOT OPTION
Select account, ISNULL([C1],0) as [C1],ISNULL([C2],0) as [C2],ISNULL([C3],0) as [C3],ISNULL([C4],0) as [C4],ISNULL([C5],0) as [C5],ISNULL([C6],0) as [C6]
from
(
SELECT account,[pmnt_number],[total_payments]
FROM payments ) AS source_tbl
PIVOT
(
MIN([Total_payments])
FOR [pmnt_number] IN ([C1],[C2],[C3],[C4],[C5],[C6])
) AS PVTTable
WITH LEFT JOIN
SELECT
rptpop.Account,
rptpop.xyz
, '$' + CONVERT (VARCHAR (12),c1.[Total_payments] , 1) as [c1]
, '$' + CONVERT (VARCHAR (12),c2.[Total_payments] , 1) as [c2]
, '$' + CONVERT (VARCHAR (12),c3.[Total_payments] , 1) as [c3]
, '$' + CONVERT (VARCHAR (12),c4.[Total_payments] , 1) as [c4]
, '$' + CONVERT (VARCHAR (12),c5.[Total_payments] , 1) as [c5]
, '$' + CONVERT (VARCHAR (12),c6.[Total_payments] , 1) as [c6]
FROM Active_Loans as rptpop
LEFT JOIN payments as c1 on c1.Account = rptpop.Account and c1.pmnt_number = 'C1'
LEFT JOIN payments as c2 on c1.Account = rptpop.Account and c1.pmnt_number = 'C2'
LEFT JOIN payments as c3 on c1.Account = rptpop.Account and c1.pmnt_number = 'C3'
LEFT JOIN payments as c4 on c1.Account = rptpop.Account and c1.pmnt_number = 'C4'
LEFT JOIN payments as c5 on c1.Account = rptpop.Account and c1.pmnt_number = 'C5'
LEFT JOIN payments as c6 on c1.Account = rptpop.Account and c1.pmnt_number = 'C6'