Related
I have a table in SQL Server with the following format
MType (Integer), MDate (Datetime), Status (SmallInt)
1, 10-05-2018, 1
1, 15-05-2018, 1
2, 25-3-2018, 0
3, 12-01-2018, 1
....
I want to get the MIN MDate for specific MTypes for future dates. In case there isn't one, then the MType should be returned but with NULL value.
Here is what I have done until now:
SELECT m.MType,
MIN(m.MDate)
FROM MyTypes m
WHERE m.MType IN ( 1, 2, 3, 4)
AND m.MDate > GETDATE()
AND m.Status = 1
GROUP BY m.MType
Obviously, the above will return only the following:
1, 10-05-2018
Since there are any other rows with future date and status equals to 1.
However, the results I want are:
1, 10-05-2018
2, NULL
3, NULL
4, NULL //this is missing in general from the table. No MType with value 4
The table is big, so performance is something to take into account. Any ideas how to proceed?
One way is to join the table to itself and filter the date in the ON clause.
SELECT a.Mtype, MIN(b.MDate)
FROM MyTypes a
LEFT JOIN MyTypes b
ON a.MType = b.MType
AND b.MDate > GETDATE()
AND b.Status = 1
WHERE a.MType IN ( 1, 2, 3)
GROUP BY a.MType
Here's a Demo.
I don't know what is logic behind but it seems to use of look-up tables
SELECT a.MType, l.MDate
FROM
(
values (1),(2),(3),(4)
)a (MType)
LEFT JOIN (
SELECT m.MType,
MIN(m.MDate) MDate
FROM MyTypes m
WHERE m.MDate > GETDATE()
AND m.Status = 1
GROUP BY m.MType
)l on l.MType = a.MType
Use a windows function and a union to a numbers table:
declare #t table (MType int, MDate datetime, [Status] smallint)
Insert into #t values (1, convert(date, '10-05-2018', 103), 1)
,(1, convert(date, '15-05-2018', 103), 1)
,(2, convert(date, '25-03-2018', 103), 0)
,(3, convert(date, '12-01-2018', 103), 1)
Select DISTINCT Mtype
, min(iiF(MDate>getdate() and status = 1, MDate, NUll)) over (Partition By Mtype) as MDate
from ( SELECT TOP 10000 row_number() over(order by t1.number) as MType
, '1900-01-01' as MDate, 0 as [Status]
FROM master..spt_values t1
CROSS JOIN master..spt_values t2
union
Select Mtype, MDate, [Status] from #t
) x
where MType in (1,2,3,4)
order by x.MType
I have attendance data list which is showing below. Now I am trying to find data by a specific date range (01/05/2016 – 07/05/2016) with total Present Column, Total Present Column will be calculated from previous present data (P). Suppose today is 04/05/2016. If a person has 01,02,03,04 status ‘p’ then it will show date 04-05-2016 total present 4.
Could you help me to find total present from this result set.
You can check this example, which have logic to calculate previous sum value.
declare #t table (employeeid int, datecol date, status varchar(2) )
insert into #t values (10001, '01-05-2016', 'P'),
(10001, '02-05-2016', 'P'),
(10001, '03-05-2016', 'P'),
(10001, '04-05-2016', 'P'),
(10001, '05-05-2016', 'A'),
(10001, '06-05-2016', 'P'),
(10001, '07-05-2016', 'P'),
(10001, '08-05-2016', 'L'),
(10002, '07-05-2016', 'P'),
(10002, '08-05-2016', 'L')
--select * from #t
select * ,
SUM(case when status = 'P' then 1 else 0 end) OVER (PARTITION BY employeeid ORDER BY employeeid, datecol
ROWS BETWEEN UNBOUNDED PRECEDING
AND current row)
from
#t
Another twist of the same thing via cte (as you written SQLSERVER2012, this below solution only work in Sqlserver 2012 and above)
;with cte as
(
select employeeid , datecol , ROW_NUMBER() over(partition by employeeid order by employeeid, datecol) rowno
from
#t where status = 'P'
)
select t.*, cte.rowno ,
case when ( isnull(cte.rowno, 0) = 0)
then LAG(cte.rowno) OVER (ORDER BY t.employeeid, t.datecol)
else cte.rowno
end LagValue
from #t t left join cte on t.employeeid = cte.employeeid and t.datecol = cte.datecol
order by t.employeeid, t.datecol
You could use a subquery to calculate TotalPresent for each row:
SELECT
main.EmployeeID,
main.[Date],
main.[Status],
(
SELECT SUM(CASE WHEN t.[Status] = 'P' THEN 1 ELSE 0 END)
FROM [TableName] t
WHERE t.EmployeeID = main.EmployeeID AND t.[Date] <= main.[Date]
) as TotalPresent
FROM [TableName] main
ORDER BY
main.EmployeeID,
main.[Date]
Here I used subquery to count the sum of records that have the same EmployeeID and date is less or equal to the date of current row. If status of the record is 'P', then 1 is added to the sum, otherwise 0, which counts only records that have status P.
Interesting question, this should work:
select *
, (select count(retail) from p g
where g.date <= p.date and g.id = p.id and retail = 'P')
from p
order by ID, Date;
So I believe I understand correctly. You would like to count the occurences of P per ID datewise.
This makes a lot of sense. That is why the first occurrence of ID2 was L and the Total is 0. This query will count P status for each occurrence, pause at non-P for each ID.
Here is an example
So I have a Visitor table, and a Visitor_activity table. Say:
Visitor
Visitor_ID Int
Visitor_name varchar(20)
Visitor_Activity
ID Int
Visitor_ID Int
Activity_Type char(3) -- values IN or OUT
Activity_Time datetime
Visitors might sign in and out multiple times in a day.
I'd like a nice query to tell me all visitors who are in: i.e. the last activity for today (on activity_time) was an "IN" not an "OUT". Any advice much appreciated.
It's T-SQL by the way, but I think it's more of an in-principle question.
One way to solve this is to use a correlated not exists predicate:
select Activity_Time, Visitor_ID
from Visitor_Activity t1
where Activity_Type = 'IN'
and not exists (
select 1
from Visitor_Activity
where Activity_Type = 'OUT'
and Visitor_ID = t1.Visitor_ID
and Activity_Time > t1.Activity_Time
and cast(Activity_Time as date) = cast(t1.Activity_Time as date)
)
This basically says get all visitor_id that have type = IN for which there doesn't exists any type = OUT record with a later time (on the same date).
Sample SQL Fiddle
SELECT
v.*
FROM
Visitors v
JOIN Visitor_Activity va ON va.Visitor_ID = v.Visitor_ID
WHERE
va.Activity_Type = 'IN'
AND NOT EXISTS ( SELECT
*
FROM
Visitor_Activity va_out
WHERE
va_out.Visitor_ID = va.Visitor_ID
AND va_out.Activity_Type = 'OUT'
AND va_out.Activity_Time > va.Activity_Time )
with visitorsInOut as (
select Visitor_id,
max(case when Activity_Type = 'in' then Activity_Time else null end) inTime,
max(case when Activity_Type = 'out' then Activity_Time else null end) outTime
from Visitor_Activity
where datediff(dd, Activity_Time, getdate()) = 0
group by Visitor_id)
select Visitor_id
from visitorsInOut
where inTime > outTime or outTime is null
This uses a CTE to find the activity record with the greatest Activity_Time where the Activity_Type = 'IN' and assigns it RowNum 1. Then you can INNER JOIN the CTE to the Visitor table, filtering by the CTE results where RowNum = 1.
; WITH VisAct AS(
SELECT act.Visitor_ID
, ROW_NUMBER() OVER(PARTITION BY Visitor_ID ORDER BY Activity_Time DESC) AS RowNum
FROM Visitor_Activity act
WHERE act.Activity_Type = 'IN'
AND act.Activity_Time >= CAST(GETDATE() AS DATE)
)
SELECT vis.Visitor_ID, vis.Visitor_name
FROM Visitor vis
INNER JOIN VisAct act
ON act.Visitor_ID = vis.Visitor_ID
WHERE act.Row_Num = 1
You can pull the most recent action for each visitor, and then only return those where the last action for today was to check in.
SELECT v.Visitor_ID, v.Visitor_Name, va.Activity_Type, va.Activity_Time
FROM Visitor AS v
INNER JOIN (SELECT Visitor_ID, Activity_Type, Activity_Time, RANK() OVER (PARTITION BY Visitor_ID ORDER BY Activity_Time DESC) AS LastAction
FROM Visitor_Activity
-- checks for today, can be omitted if you still want
-- to see someone checked in from yesterday
WHERE DATEDIFF(d, 0, Activity_Time) = DATEDIFF(d, 0, getdate())
) AS va ON va.Visitor_ID = v.Visitor_ID
WHERE LastAction = 1
AND Activity_Type = 'IN'
With CROSS APPLY:
DECLARE #d DATE = '20150320'
DECLARE #v TABLE
(
visitor_id INT ,
visitor_name NVARCHAR(MAX)
)
DECLARE #a TABLE
(
visitor_id INT ,
type CHAR(3) ,
time DATETIME
)
INSERT INTO #v
VALUES ( 1, 'A' ),
( 2, 'B' ),
( 3, 'C' )
INSERT INTO #a
VALUES ( 1, 'in', '2015-03-20 19:32:27.513' ),
( 1, 'out', '2015-03-20 19:32:27.514' ),
( 1, 'in', '2015-03-20 19:32:27.515' ),
( 2, 'in', '2015-03-20 19:32:27.516' ),
( 2, 'out', '2015-03-20 19:32:27.517' ),
( 3, 'in', '2015-03-20 19:32:27.518' ),
( 3, 'out', '2015-03-20 19:32:27.519' ),
( 3, 'in', '2015-03-20 19:32:27.523' )
SELECT *
FROM #v v
CROSS APPLY ( SELECT *
FROM ( SELECT TOP 1
type
FROM #a a
WHERE a.visitor_id = v.visitor_id
AND a.time >= #d
AND a.time < DATEADD(dd, 1, #d)
ORDER BY time DESC
) i
WHERE type = 'in'
) c
Output:
visitor_id visitor_name type
1 A in
3 C in
The principle:
First you are selecting all visitors.
Then you are applying to visitor last activity
SELECT TOP 1
type
FROM #a a
WHERE a.visitor_id = v.visitor_id
AND a.time >= #d
AND a.time < DATEADD(dd, 1, #d)
ORDER BY time DESC
Then you are selecting from previous step in order to get empty set which will filter out visitors whose last activity was not 'in'. If last activity was 'in' you get one row in result and thus applying works. If last activity is 'out' then outer query will result in empty set, and by design CROSS APPLY will eliminate such visitor.
I have hospital patient admission data in Microsoft SQL Server r2 that looks something like this:
PatientID, AdmitDate, DischargeDate
Jones. 1-jan-13 01:37. 1-jan-13 17:45
Smith 1-jan-13 02:12. 2-jan-13 02:14
Brooks. 4-jan-13 13:54. 5-jan-13 06:14
I would like count the number of patients in the hospital day by day and hour by hour (ie at
1-jan-13 00:00. 0
1-jan-13 01:00. 0
1-jan-13 02:00. 1
1-jan-13 03:00. 2
And I need to include the hours when there are no patients admitted in the result.
I can't create tables so making a reference table listing all the hours and days is out, though.
Any suggestions?
To solve this problem, you need a list of date-hours. The following gets this from the admit date cross joined to a table with 24 hours. The table of 24 hours is calculating from information_schema.columns -- a trick for getting small sequences of numbers in SQL Server.
The rest is just a join between this table and the hours. This version counts the patients at the hour, so someone admitted and discharged in the same hour, for instance is not counted. And in general someone is not counted until the next hour after they are admitted:
with dh as (
select DATEADD(hour, seqnum - 1, thedatehour ) as DateHour
from (select distinct cast(cast(AdmitDate as DATE) as datetime) as thedatehour
from Admission a
) a cross join
(select ROW_NUMBER() over (order by (select NULL)) as seqnum
from INFORMATION_SCHEMA.COLUMNS
) hours
where hours <= 24
)
select dh.DateHour, COUNT(*) as NumPatients
from dh join
Admissions a
on dh.DateHour between a.AdmitDate and a.DischargeDate
group by dh.DateHour
order by 1
This also assumes that there are admissions on every day. That seems like a reasonable assumption. If not, a calendar table would be a big help.
Here is one (ugly) way:
;WITH DayHours AS
(
SELECT 0 DayHour
UNION ALL
SELECT DayHour+1
FROM DayHours
WHERE DayHour+1 <= 23
)
SELECT B.AdmitDate, A.DayHour, COUNT(DISTINCT PatientID) Patients
FROM DayHours A
CROSS JOIN (SELECT DISTINCT CONVERT(DATE,AdmitDate) AdmitDate
FROM YourTable) B
LEFT JOIN YourTable C
ON B.AdmitDate = CONVERT(DATE,C.AdmitDate)
AND A.DayHour = DATEPART(HOUR,C.AdmitDate)
GROUP BY B.AdmitDate, A.DayHour
This is a bit messy and includes a temp table with the test data you provided but
CREATE TABLE #HospitalPatientData (PatientId NVARCHAR(MAX), AdmitDate DATETIME, DischargeDate DATETIME)
INSERT INTO #HospitalPatientData
SELECT 'Jones.', '1-jan-13 01:37:00.000', '1-jan-13 17:45:00.000' UNION
SELECT 'Smith', '1-jan-13 02:12:00.000', '2-jan-13 02:14:00.000' UNION
SELECT 'Brooks.', '4-jan-13 13:54:00.000', '5-jan-13 06:14:00.000'
;WITH DayHours AS
(
SELECT 0 DayHour
UNION ALL
SELECT DayHour+1
FROM DayHours
WHERE DayHour+1 <= 23
),
HospitalPatientData AS
(
SELECT CONVERT(nvarchar(max),AdmitDate,103) as AdmitDate ,DATEPART(hour,(AdmitDate)) as AdmitHour, COUNT(PatientID) as CountOfPatients
FROM #HospitalPatientData
GROUP BY CONVERT(nvarchar(max),AdmitDate,103), DATEPART(hour,(AdmitDate))
),
Results AS
(
SELECT MAX(h.AdmitDate) as Date, d.DayHour
FROM HospitalPatientData h
INNER JOIN DayHours d ON d.DayHour=d.DayHour
GROUP BY AdmitDate, CountOfPatients, DayHour
)
SELECT r.*, COUNT(h.PatientId) as CountOfPatients
FROM Results r
LEFT JOIN #HospitalPatientData h ON CONVERT(nvarchar(max),AdmitDate,103)=r.Date AND DATEPART(HOUR,h.AdmitDate)=r.DayHour
GROUP BY r.Date, r.DayHour
ORDER BY r.Date, r.DayHour
DROP TABLE #HospitalPatientData
This may get you started:
BEGIN TRAN
DECLARE #pt TABLE
(
PatientID VARCHAR(10)
, AdmitDate DATETIME
, DischargeDate DATETIME
)
INSERT INTO #pt
( PatientID, AdmitDate, DischargeDate )
VALUES ( 'Jones', '1-jan-13 01:37', '1-jan-13 17:45' ),
( 'Smith', '1-jan-13 02:12', '2-jan-13 02:14' )
, ( 'Brooks', '4-jan-13 13:54', '5-jan-13 06:14' )
DECLARE #StartDate DATETIME = '20130101'
, #FutureDays INT = 7
;
WITH dy
AS ( SELECT TOP (#FutureDays)
ROW_NUMBER() OVER ( ORDER BY name ) dy
FROM sys.columns c
) ,
hr
AS ( SELECT TOP 24
ROW_NUMBER() OVER ( ORDER BY name ) hr
FROM sys.columns c
)
SELECT refDate, COUNT(p.PatientID) AS PtCount
FROM ( SELECT DATEADD(HOUR, hr.hr - 1,
DATEADD(DAY, dy.dy - 1, #StartDate)) AS refDate
FROM dy
CROSS JOIN hr
) ref
LEFT JOIN #pt p ON ref.refDate BETWEEN p.AdmitDate AND p.DischargeDate
GROUP BY refDate
ORDER BY refDate
ROLLBACK
Suppose I have the following an event table with personId, startDate and endDate.
I want to know how much time the person X spent doing an event (the events can override each other).
If the person just has 1 event, its easy: datediff(dd, startDate, endDate)
If the person has 2 events it gets tricky.
I'll set some scenarios for the expected results.
Scenario 1
startDate endDate
1 4
3 5
This means he the results should be the datediff from 1 to 5
Scenario 2
startDate endDate
1 3
6 9
this means he the results should be the some of datediff(dd,1,3) and datediff(dd,6,9)
How can I get this result on an sql query? I can only think of a bunch of if statements, but the same person can have n events so the query will be really confusing.
Shredder Edit: I'd like to add a 3rd scenario:
startDate endDate
1 5
4 8
11 15
Desired result to Shredder scenario:
(1,5) and (4,8) merge in (1,8) since they overlap then we need to datediff(1,8) + datediff(11,15) => 7 + 4 => 11
You can use a recursive CTE to build a list of dates and then count the distinct dates.
declare #T table
(
startDate date,
endDate date
);
insert into #T values
('2011-01-01', '2011-01-05'),
('2011-01-04', '2011-01-08'),
('2011-01-11', '2011-01-15');
with C as
(
select startDate,
endDate
from #T
union all
select dateadd(day, 1, startDate),
endDate
from C
where dateadd(day, 1, startDate) < endDate
)
select count(distinct startDate) as DayCount
from C
option (MAXRECURSION 0)
Result:
DayCount
-----------
11
Or you can use a numbers table. Here I use master..spt_values:
declare #MinStartDate date
select #MinStartDate = min(startDate)
from #T
select count(distinct N.number)
from #T as T
inner join master..spt_values as N
on dateadd(day, N.Number, #MinStartDate) between T.startDate and dateadd(day, -1, T.endDate)
where N.type = 'P'
Here's a solution that uses the Tally table idea (which I first heard of in an article by Itzk Ben-Gan -- I still cut and paste his code whenver the subject comes up). The idea is to generate a list of ascending integers, join the source data by range against the numbers, and then count the number of distinct numbers, as follows. (This code uses syntax from SQL Server 2008, but with minor modifications would work in SQL 2005.)
First set up some testing data:
CREATE TABLE #EventTable
(
PersonId int not null
,startDate datetime not null
,endDate datetime not null
)
INSERT #EventTable
values (1, 'Jan 1, 2011', 'Jan 4, 2011')
,(1, 'Jan 3, 2011', 'Jan 5, 2011')
,(2, 'Jan 1, 2011', 'Jan 3, 2011')
,(2, 'Jan 6, 2011', 'Jan 9, 2011')
Determine some initial values
DECLARE
#Interval bigint
,#FirstDay datetime
,#PersonId int = 1 -- (or whatever)
Get the first day and the maximum possible number of dates (to keep the cte from generating extra values):
SELECT
#Interval = datediff(dd, min(startDate), max(endDate)) + 1
,#FirstDay = min(startDate)
from #EventTable
where PersonId = #PersonId
Cut and paste over the one routine and modify and test it to only return as many integers as we'll need:
/*
;WITH
Pass0 as (select 1 as C union all select 1), --2 rows
Pass1 as (select 1 as C from Pass0 as A, Pass0 as B),--4 rows
Pass2 as (select 1 as C from Pass1 as A, Pass1 as B),--16 rows
Pass3 as (select 1 as C from Pass2 as A, Pass2 as B),--256 rows
Pass4 as (select 1 as C from Pass3 as A, Pass3 as B),--65536 rows
Pass5 as (select 1 as C from Pass4 as A, Pass4 as B),--4,294,967,296 rows
Tally as (select row_number() over(order by C) as Number from Pass5)
select Number from Tally where Number <= #Interval
*/
And now revise it by first joining to the intervals defined in each source row, and then count each distinct value found:
;WITH
Pass0 as (select 1 as C union all select 1), --2 rows
Pass1 as (select 1 as C from Pass0 as A, Pass0 as B),--4 rows
Pass2 as (select 1 as C from Pass1 as A, Pass1 as B),--16 rows
Pass3 as (select 1 as C from Pass2 as A, Pass2 as B),--256 rows
Pass4 as (select 1 as C from Pass3 as A, Pass3 as B),--65536 rows
Pass5 as (select 1 as C from Pass4 as A, Pass4 as B),--4,294,967,296 rows
Tally as (select row_number() over(order by C) as Number from Pass5)
SELECT PersonId, count(distinct Number) EventDays
from #EventTable et
inner join Tally
on dateadd(dd, Tally.Number - 1, #FirstDay) between et.startDate and et.endDate
where et.PersonId = #PersonId
and Number <= #Interval
group by PersonId
Take out the #PersonId filter and you'd get it for all persons. And with minor modification you can do it for any time interval, not just days (which is why I set the Tally table to generate severely large numbers.)
The following SQL is for the three scenarios you've described
with sampleData
AS (
SELECT 1 personid,1 startDate,4 endDate
UNION SELECT 1,3,5
UNION SELECT 2,1,3
UNION SELECT 2,6,9
UNION SELECT 3,1,5
UNION SELECT 3,4,8
UNION SELECT 3,11, 15
),
cte
AS (SELECT personid,
startdate,
enddate,
Row_number() OVER(ORDER BY personid, startdate) AS rn
FROM sampledata),
overlaps
AS (SELECT a.personid,
a.startdate,
b.enddate,
a.rn id1,
b.rn id2
FROM cte a
INNER JOIN cte b
ON a.personid = b.personid
AND a.enddate > b.startdate
AND a.rn = b.rn - 1),
nooverlaps
AS (SELECT a.personid,
a.startdate,
a.enddate
FROM cte a
LEFT JOIN overlaps b
ON a.rn = b.id1
OR a.rn = b.id2
WHERE b.id1 IS NULL)
SELECT personid,
SUM(timespent) timespent
FROM (SELECT personid,
enddate - startdate timespent
FROM nooverlaps
UNION
SELECT personid,
enddate - startdate
FROM overlaps) t
GROUP BY personid
Produces this result
Personid timeSpent
----------- -----------
1 4
2 5
3 11
Notes: I used the simple integers but the DateDiffs should work too
Correctness issue There is a correctness issue if your data is allowed to have multiple overlaps as Cheran S noted, the results won't be correct and you should use one of the other answers instead. His example used [1,5],[4,8],[7,11] for the same person ID
Algebra. If B-n is the ending time of the nth event, and A-n is the starting time of the nth event, then the sum of the differences is the difference of the sums. So you can write
select everything else, sum(cast(endDate as int)) - sum(cast(startDate as int)) as daysSpent
If your dates have no time component, this works. Otherwise, you could use a real.
Try something like this
select
personId,
sum(DateDuration) as TotalDuration
from
(
select personId, datediff(dd, startDate, endDate) as DateDuration
from yourEventTable
) a
group by personId
;WITH cte(gap)
AS
(
SELECT sum(b-a) from xxx GROUP BY uid
)
SELECT * FROM cte
Edit 1: I have modified both solutions to get correct results.
Edit 2: I have done comparative tests using the solutions proposed by Mikael Eriksson, Conrad Frix, Philip Kelley and me. All tests use an EventTable with the following structure:
CREATE TABLE EventTable
(
EventID INT IDENTITY PRIMARY KEY
,PersonId INT NOT NULL
,StartDate DATETIME NOT NULL
,EndDate DATETIME NOT NULL
,CONSTRAINT CK_StartDate_Before_EndDate CHECK(StartDate < EndDate)
);
Also, all tests use warm buffer (no DBCC DROPCLEANBUFFERS) and cold [plan] cache (I have executed DBCC FREEPROCCACHE before every test). Because some solutions use a filter(PersonId = 1) and others not, I have inserted into EventTable rows for only one person (INSERT ...(PersonId,...) VALUES (1,...)).
These are the results:
My solutions use recursive CTEs.
Solution 1:
WITH BaseCTE
AS
(
SELECT e.StartDate
,e.EndDate
,e.PersonId
,ROW_NUMBER() OVER(PARTITION BY e.PersonId ORDER BY e.StartDate, e.EndDate) RowNumber
FROM EventTable e
), RecursiveCTE
AS
(
SELECT b.PersonId
,b.RowNumber
,b.StartDate
,b.EndDate
,b.EndDate AS MaxEndDate
,1 AS PseudoDenseRank
FROM BaseCTE b
WHERE b.RowNumber = 1
UNION ALL
SELECT crt.PersonId
,crt.RowNumber
,crt.StartDate
,crt.EndDate
,CASE WHEN crt.EndDate > prev.MaxEndDate THEN crt.EndDate ELSE prev.MaxEndDate END
,CASE WHEN crt.StartDate <= prev.MaxEndDate THEN prev.PseudoDenseRank ELSE prev.PseudoDenseRank + 1 END
FROM RecursiveCTE prev
INNER JOIN BaseCTE crt ON prev.PersonId = crt.PersonId
AND prev.RowNumber + 1 = crt.RowNumber
), SumDaysPerPersonAndInterval
AS
(
SELECT src.PersonId
,src.PseudoDenseRank --Interval ID
,DATEDIFF(DAY, MIN(src.StartDate), MAX(src.EndDate)) Days
FROM RecursiveCTE src
GROUP BY src.PersonId, src.PseudoDenseRank
)
SELECT x.PersonId, SUM( x.Days ) DaysPerPerson
FROM SumDaysPerPersonAndInterval x
GROUP BY x.PersonId
OPTION(MAXRECURSION 32767);
Solution 2:
DECLARE #Base TABLE --or a temporary table: CREATE TABLE #Base (...)
(
PersonID INT NOT NULL
,StartDate DATETIME NOT NULL
,EndDate DATETIME NOT NULL
,RowNumber INT NOT NULL
,PRIMARY KEY(PersonID, RowNumber)
);
INSERT #Base (PersonID, StartDate, EndDate, RowNumber)
SELECT e.PersonId
,e.StartDate
,e.EndDate
,ROW_NUMBER() OVER(PARTITION BY e.PersonID ORDER BY e.StartDate, e.EndDate) RowNumber
FROM EventTable e;
WITH RecursiveCTE
AS
(
SELECT b.PersonId
,b.RowNumber
,b.StartDate
,b.EndDate
,b.EndDate AS MaxEndDate
,1 AS PseudoDenseRank
FROM #Base b
WHERE b.RowNumber = 1
UNION ALL
SELECT crt.PersonId
,crt.RowNumber
,crt.StartDate
,crt.EndDate
,CASE WHEN crt.EndDate > prev.MaxEndDate THEN crt.EndDate ELSE prev.MaxEndDate END
,CASE WHEN crt.StartDate <= prev.MaxEndDate THEN prev.PseudoDenseRank ELSE prev.PseudoDenseRank + 1 END
FROM RecursiveCTE prev
INNER JOIN #Base crt ON prev.PersonId = crt.PersonId
AND prev.RowNumber + 1 = crt.RowNumber
), SumDaysPerPersonAndInterval
AS
(
SELECT src.PersonId
,src.PseudoDenseRank --Interval ID
,DATEDIFF(DAY, MIN(src.StartDate), MAX(src.EndDate)) Days
FROM RecursiveCTE src
GROUP BY src.PersonId, src.PseudoDenseRank
)
SELECT x.PersonId, SUM( x.Days ) DaysPerPerson
FROM SumDaysPerPersonAndInterval x
GROUP BY x.PersonId
OPTION(MAXRECURSION 32767);