how to have column output zero when it equals null? - sql

Question

I have a query where i get the average of some ratings
which works correct.
WITH row_avg_table(avg_rating,employee, approveddate) AS
(SELECT
(SELECT AVG(rating)
FROM (
VALUES (CAST(c.rating1 AS float)), (CAST(c.rating2 AS float)), (CAST(c.rating3 AS float)),
(CAST(c.rating4 AS float)), (CAST(c.rating5 AS float)) ) AS v (rating)
WHERE v.rating > 0) avg_rating,
employee,approveddate
FROM CSEReduxResponses c)
SELECT employee,
avg(avg_rating) as average_rating
FROM row_avg_table
where month(approveddate)=2014
AND year(approveddate)=6
GROUP BY employee;
The problem im having is when a rating 1-5 would all be 0.
Right now it gives me 'null' i would like it to show 0 for this special occasion.
for example i have the data below
create table CSEReduxResponses (rating1 int, rating2 int, rating3 int, rating4 int, rating5 int,
approveddate datetime,employee int)
insert into CSEReduxResponses (rating1 , rating2 ,rating3 , rating4 , rating5 ,
approveddate, employee )
values
(5,4,5,1,4,'2014-06-18',1),
(5,4,5,1,0,'2014-06-18',1),
(0,0,0,0,0,'2014-06-19',3);
So for employee=3 average_rating =0

Answer 1

What if you use ISNULL function like below
SELECT ISNULL(AVG(rating),some_default_value)
(OR) CASE condition like below
SELECT AVG(case when rating = 0 then 1 else rating end)

Answer 2

The problem you have is the WHERE v.rating > 0 clause. If the ratings are 0, then this will return no values - and the average of nothing cannot be calculated, and thus returns null. Simply remove this clause, or if you want to filter negative values, change it to WHERE v.rating >= 0.
If you use 0 to store values that you don't want to impact the rating (like it appears that you might be doing from the second of the examples you provide) this won't work - you'll have to instead go through the output and replace null values with 0. You might copy the results from your first query into a temp table/table variable and do replace such as
select employee, rating = isnull(t.rating, 0)
from #temptable t
Or something of the like.

Answer 3

Use
isnull(Rating1, 0)
or
coalesce(Rating1, 0)
If Rating1 is null then it will be replaced with the default value 0. Similartly you can check for other fields. The COALESCE and ISNULL T-SQL functions are used to return the first nonnull expression among the input arguments but there are few differences between those which you can find here
More details on coalesce

Answer 4

There is no advantage in using a CTE for this. You are unpivoting the non-normalized table using cross apply and values. To arrive at the equivalent of the unpivot command you use WHERE ... IS NOT NULL but presumably you are ensuring all 5 values are present to avoid bias in the calculations, so I would suggest this:
SELECT
employee
, AVG(isnull(rating,0)) AS avg_rating
FROM CSEReduxResponses c
CROSS APPLY (
VALUES (CAST(c.rating1 AS float))
, (CAST(c.rating2 AS float))
, (CAST(c.rating3 AS float))
, (CAST(c.rating4 AS float))
, (CAST(c.rating5 AS float))
) AS ca1 (rating)
WHERE approveddate >= '20140601'
AND approveddate < '20140701'
GROUP BY
employee
;
results from your sample data:
| EMPLOYEE | AVG_RATING |
|----------|------------|
| 1 | 3.4 |
| 3 | 0 |
see: http://sqlfiddle.com/#!3/f91d9/2
Also: I would encourage you to NOT to use functions on the data to facilitate a where condition. Avoiding functions on data allows use of indexes. Here all you need is a simple date range (I have used YYYYMMDD as it is the safest format for sql server).