SQL Server counting number of rows from beginning of sorted table where colvalue = 0

Question

We have SQL Server 2008 R2. I have table MyTable .

I need:

1. sort by column Nr (ORDER BY MyTable.NR)

2. count how many values are equal to 0 from the beginning of this sorted table (taking in mind that it is already sorted by NR) ( MyTable.Value = 0 ).

For instance, if MyTable has these values:

NR Value
2  0
1  0
3  5
4  0

then I have to get count = 2, because if sort this table by Nr we have two rows where Value = 0 from the beginning of the table.

Answer 1

You can do this using NOT EXISTS to exclude all rows where a row exists with a lower NR and a Value <> 0. eg

DECLARE @T TABLE (NR INT, Value INT);
INSERT @T VALUES (2, 0), (1, 0), (3, 5), (4, 0);

SELECT  COUNT(*)
FROM    @T AS t1
WHERE   t1.Value = 0
AND     NOT EXISTS
        (   SELECT  1
            FROM    @T AS t2
            WHERE   t2.Value <> 0
            AND     t2.NR < t1.NR
        );

Answer 2

You can filter data using predicate Value = 0 , then number filtered data with row_number , and then count rows, where assigned number equal to original one:

declare @data table (NR int, Value int);
insert into @data values
    (2, 0), (1, 0), (3, 5), (4, 0);

;with
step1 as (select NR from @data where Value = 0),
step2 as (select NR, rn = row_number() over (order by NR) from step1)
select count(1)
from step2
where NR = rn;