Min Max (Date Time) And Count Data

Question

Using SQL Server 2014

Example Table:

|   ID  |   Date_Time               |   ID_Plan |
|   1   |   2018-12-19 21:21:00.000 |   A1      |
|   2   |   2018-12-19 21:22:00.000 |   A1      |
|   3   |   2018-12-19 21:23:00.000 |   NULL    |
|   4   |   2018-12-19 21:24:00.000 |   NULL    |
|   5   |   2018-12-19 21:25:00.000 |   A2      |
|   6   |   2018-12-19 21:26:00.000 |   A2      |
|   7   |   2018-12-19 21:27:00.000 |   A2      |
|   8   |   2018-12-19 21:28:00.000 |   A3      |
|   9   |   2018-12-19 21:29:00.000 |   A3      |
|   10  |   2018-12-19 21:30:00.000 |   NULL    |
|   11  |   2018-12-19 21:31:00.000 |   NULL    |
|   12  |   2018-12-19 21:32:00.000 |   NULL    |
|   13  |   2018-12-19 21:33:00.000 |   A4      |
|   14  |   2018-12-19 21:34:00.000 |   A4      |
|   15  |   2018-12-19 21:35:00.000 |   A4      |
|   16  |   2018-12-20 21:36:00.000 |   NULL    |

IF OBJECT_ID('tempdb..#tb_CountFinal') IS NOT NULL
DROP TABLE #tb_CountFinal

CREATE TABLE #tb_CountFinal
(
    [ID] [int] IDENTITY(1,1) NOT NULL,
    Date_Time DATETIME,
    ID_Plan NVARCHAR(100)
)


INSERT INTO #tb_CountFinal
SELECT '12/19/2018 21:21','A1'union all
SELECT '12/19/2018 21:22','A1'union all
SELECT '12/19/2018 21:23',NULL union all
SELECT '12/19/2018 21:24',NULL union all
SELECT '12/19/2018 21:25','A2'union all
SELECT '12/19/2018 21:26','A2'union all
SELECT '12/19/2018 21:27','A2'union all
SELECT '12/19/2018 21:28','A3'union all
SELECT '12/19/2018 21:29','A3'union all
SELECT '12/19/2018 21:30',NULL union all
SELECT '12/19/2018 21:31',NULL union all
SELECT '12/19/2018 21:32',NULL union all
SELECT '12/19/2018 21:33','A4'union all
SELECT '12/19/2018 21:34','A4'union all
SELECT '12/19/2018 21:35','A4'union all
SELECT '12/20/2018 21:36',NULL

I'm interested in a query which would output rows similar to the following:

|   Start_Date          |       End_Date        |   Plan_ID     |   Count_PlantID   |
|   12/19/2018 21:21    |   12/19/2018 21:22    |   A1          |   2               |
|   12/19/2018 21:23    |   12/19/2018 21:24    |   NULL        |   2               |
|   12/19/2018 21:25    |   12/19/2018 21:27    |   A2          |   3               |
|   12/19/2018 21:28    |   12/19/2018 21:29    |   A3          |   2               |
|   12/19/2018 21:30    |   12/19/2018 21:32    |   NULL        |   3               |
|   12/19/2018 21:33    |   12/19/2018 21:35    |   A4          |   3               |
|   12/20/2018 21:36    |   12/20/2018 21:36    |   NULL        |   1               |

Answer 1

For each NULL entries I made a key by looking at the previous starting [ID] column, then performed the Group By on that new column. Try it.

select  min(Date_Time) as Start_Date, max(Date_Time) as End_Date
        , min(ID_Plan) as Plan_ID, count(1) as Count_PlanID
from    (
        select  s.*
                , case when ID_Plan is null 
                    then 
                        cast(
                                (
                                select  max([ID]) 
                                from    #tb_CountFinal as x 
                                where x.ID < s.ID and ID_Plan is not null
                                ) as varchar(100)
                            )
                    else 
                        ID_Plan 
                    end as ID_Plan_Adv_Start
        from    #tb_CountFinal  as s
        ) as dt
group by dt.ID_Plan_Adv_Start
order by Start_Date