Add zeros to a matrix report

Question

I'm making a matrix report and the users want to have zeros show where is no data returned. Since there are no fields I can't set the null values to zero. I used the example in the building reports pdf but it covers up the fields that returns numbers. Then I noticed the example bascially just adds the zeros to the end since they are using currency. Is there a way to have a label moveforward only when you want it, or is there another way that I'm missing.

Answer 1

OK – I can now see that I wasn’t clear enough in my first reply. I have modified your SQL statement (hopefully without syntax errors :-) The job was to get all possible/necessary combinations of EVENT and SEVERITY out of sub query SUBALLROWS (i.e. I only changed sub query SUBALLROWS):
[All my indentions are gone when I Post the Message, so it looks horrible.]
SELECT suballrows.event,
suballrows.severity,
DECODE(suballrows.severity,'R',1,'A',2,'G',3) AS SEVERITYSORT,
nvl(count(subnew.severity),0) as SEVERITYTOTAL
FROM (SELECT severity,
event
FROM vw_merger_drill_issue
WHERE bau_merger = '1'
AND trunc(current_date-submitdate)<7
)subnew,
(SELECT event, -- Cartesian product of event and severity
severity
FROM (SELECT DISTINCT event
FROM vw_merger_drill_issue
WHERE event IS NOT NULL
) subevent,
(SELECT DISTINCT severity
FROM vw_merger_drill_issue
WHERE severity IS NOT NULL
) subseverity
) suballrows
WHERE suballrows.event = subnew.event (+)
AND suballrows.severity = subnew.severity (+)
GROUP BY suballrows.event,
suballrows.severity
ORDER BY Suballrows.Event,
severitysort;
Because you more or less execute the same SELECT 3 times, you might get a performance gain by the use of the WITH clause.
My previously used detail table SALES (the same as your vw_merger_drill_issue view) contains:
SELECT *
FROM sales;
MO REGION AMOUNT
01 EUROPE 100
02 EUROPE 50
02 AMERICA 200
03 AMERICA 150
The With clause SQL looks like (and returns every month and region):
WITH
subsales AS
(SELECT month, region, sum( amount) amount
FROM sales
GROUP BY month, region
SELECT suballrows.month,
suballrows.region,
nvl( subsales.amount, 0) amount
FROM subsales,
(SELECT month, region -- Cartesian product with all possible combinations.
FROM (SELECT DISTINCT region
FROM subsales
) subregion,
(SELECT DISTINCT month
FROM subsales
) submonth
) suballrows
WHERE suballrows.month = subsales.month (+)
AND suballrows.region = subsales.region (+)
ORDER BY suballrows.month,
suballrows.region;
MO REGION AMOUNT
01 AMERICA 0
01 EUROPE 100
02 AMERICA 200
02 EUROPE 50
03 AMERICA 150
03 EUROPE 0
What your SQL would look like ... something like below. You may decide if it looks OK:
WITH
sub_merger_drill_issue AS
(SELECT event,
severity,
count(*) severity_subtotal
FROM vw_merger_drill_issue
WHERE bau_merger = '1'
AND trunc(current_date-submitdate)<7
AND event IS NOT NULL
AND severity IS NOT NULL
SELECT suballrows.event,
suballrows.severity,
DECODE(suballrows.severity,'R',1,'A',2,'G',3) AS SEVERITYSORT,
nvl(SUM(subnew.severity),0) as SEVERITYTOTAL -- count changed to sum because of sub total count in sub_merger_drill_issue.
FROM sub_merger_drill_issue subnew,
(SELECT event, -- Cartesian product of event and severity
severity
FROM (SELECT DISTINCT event
FROM sub_merger_drill_issue
) subevent,
(SELECT DISTINCT severity
FROM sub_merger_drill_issue
) subseverity
) suballrows
WHERE suballrows.event = subnew.event (+)
AND suballrows.severity = subnew.severity (+)
GROUP BY suballrows.event,
suballrows.severity
ORDER BY Suballrows.Event,
severitysort;