sql how to get consecutive appearance of value - sql

Question

suppose I have a column 'value', which can appear multiple times in a table with another column 'result' which can be either 1 or 0. I would like to search for consecutive 1s (ie result = 1) until the count reaches 4, then I can select value. given the result sets below:
-result set a)
value Result
----- ------
A 1
A 1
A 1
A 0
-result set b)
value Result
----- ------
A 1
A 1
A 1
A 1
result set b meets the condition and therefore value A is selected. How do I go about this ? Thanks.

Answer 1

This is the query: (usually this query is to detect double record in a table, but probably meet your demand).
select value, result, count(value) as [Result Sum]
from #temp
where result = 1
group by value, result
having count(value) >3
This is the Result
value result Result Sum
----- ----------- -----------
A 1 4
UPDATED:
This is the data example in my temporary table (#temp)
value result
----- -----------
A 1
A 1
A 1
A 0
A 1
D 1
D 1
D 1
D 1
B 1
B 1
C 1
C 1
C 1
C 1
From The example data C and D are the valid values
Declare #temp2 table
(
value nvarchar(5)
)
declare #value nvarchar(5), #result int, #total int, #flag bit, #tempValue nvarchar(5)
DECLARE myCursor CURSOR FOR
SELECT value, result
FROM #temp
set #flag = 1
set #tempValue = ''
OPEN myCursor;
FETCH NEXT FROM myCursor into #value, #result;
WHILE ##FETCH_STATUS = 0
BEGIN
--logic here
if (#tempValue <> #value and #result = 1) or #flag = 1
begin
set #tempValue = #value
set #total = 1
set #flag = 0
end
else --#tempvalue = #value
begin
if #result = 1
set #total = #total + 1
else --#result = 0
set #flag = 1
if #total >3 --valid value has reached 4 consecutive result =1
begin
set #flag = 1
insert into #temp2 values (#value)
end
end
FETCH NEXT FROM myCursor into #value, #result;
END;
CLOSE myCursor;
DEALLOCATE myCursor;
select * from #temp2
This is the Result of the loop (table #temp2)
value
-----
D
C
(2 row(s) affected)

Answer 2

You can do this in a select statement. You can find groups of items in a row by using row_number() assuming you have an id. SQL tables are inherently unordered, so you need an id or creation date or something to specify the ordering. Here is the SQL:
select value
from (select t,
(row_number() over (partition by value order by id) -
row_number() over (partition by value, results order by id)
) as grp
from table t
) t
group by value, result, grp
having count(*) > 3 and result = 1;