How to count column values in table row? - sql-server

Question

In SQL Server, i have one table in that there 15 columns are there, the columns contain the data SL,PL,CL or 8,4 like that.
empid D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15
==================================================================
sam PL 8 PL 4 8 SL 8 SL 4 CL 8 CL 8 8 8
sunny 8 CL 4 CL SL 8 4 SL 4 PL 8 8 8 8 8
zimmy 4 4 4 8 8 8 4 4 8 8 8 8 8 8 8
Like the above my table is there , now i want to calculate PL or CL or SL of every empid, please help me.

Answer 1

The need to aggregate across columns in a single row is a good indicator you're storing the data wrong. I would change it to the following:
empid D_number val
sam 1 PL
sam 2 8
sam 3 PL
In which case you can just do a count or conditional sum. For example, a conditional sum to get both PL and SL counts in one go:
select empid, sum(case when val = 'PL' then 1 else 0 end),
sum(case when val = 'SL' then 1 else 0 end)
from myTable
group by empid
Or if you're just interested in the PL count, you can simply COUNT:
select empid, count(*)
from myTable
where val = 'PL'
group by empid
However, assuming there's nothing you can do about your table schema, you have two options:
Conditional sum across each column explicitly (boy does this get ugly fast):
select empid,
case when d1 = 'PL' then 1 else 0 end +
case when d2 = 'PL' then 1 else 0 end +
case when d3 = 'PL' then 1 else 0 end +
...,
case when d1 = 'SL' then 1 else 0 end +
case when d2 = 'SL' then 1 else 0 end +
case when d3 = 'SL' then 1 else 0 end +
...
from myTable
UNPIVOT the data to make it look like the form above, then use a conditional sum:
with cte as (
select empid, D_number, val
from (
select empid, d1, d2, d3, d4, d5, ...
from myTable
) x
unpivot (val for D_number in
(d1, d2, d3, d4, d5, ...)
)
)
select empid, sum(case when val = 'PL' then 1 else 0 end),
sum(case when val = 'SL' then 1 else 0 end)
from cte
group by empid

Answer 2

Use UNPIVOT
Test data and table:
DECLARE #t table
(empid varchar(10), D1 char(2), D2 char(2), D3 char(2), D4 char(2),D5 char(2),
D6 char(2),D7 char(2), D8 char(2), D9 char(2),D10 char(2), D11 char(2),
D12 char(2), D13 char(2), D14 char(2), D15 char(2))
INSERT #t values
('sam', 'PL','8','PL','4','8','SL','8','SL','4','CL','8','CL','8','8','8'),
('sunny','8','CL','4','CL','SL','8','4','SL','4','PL','8','8','8','8','8'),
('zimmy','4','4','4','8','8','8','4','4','8','8','8','8','8','8','8')
Query:
SELECT empid, value, count(*) count
FROM #t as p
UNPIVOT
(value FOR col IN
([D1],[D2],[D3],[D4],[D5],[D6],[D7],[D8],[D9],
[D10],[D11],[D12],[D13],[D14],[D15]) ) AS unpvt
WHERE value in ('PL','CL','SL')
GROUP BY empid, value
Result:
empid value count
sam CL 2
sam PL 2
sam SL 2
sunny CL 2
sunny PL 1
sunny SL 2