Writing single query with conflicting WHERE statements

Question

How do I run the following as a single query that will result in a column that pertains to the first block of code and then another column that pertains to the 2nd block of code? The issue is that I need 2 different WHERE statements. So the final output will be patient_id, complete, incomplete.
SELECT patient_id, COUNT(*) AS complete
FROM STATUS
WHERE status = 1
GROUP BY patient_id
ORDER BY patient_id;
SELECT patient_id, COUNT(*) AS incomplete
FROM STATUS
WHERE status = 2
GROUP BY patient_id
ORDER BY patient_id;
Thanks!

Answer 1

Hi,
apex wrote:
Thanks for all of the help.
I would like to add another column with decile and am struggling as to how to do it. Since I can't reference something in a calculation in the same step, I think I will need a 3rd nesting, but what I haven't figured out is how to get the number of subjects whose ratio is less than that subject's ratio. Right: you can't assign an alias (such as complete or ratio) to a calculated column and use that alias in the same sub-query.
If the calculation isn't very complicated, then you might find it simpler just to repeat the calculation. For example, I think this is what you want:
WITH     got_complete     AS
     SELECT    patient_id
     ,        COUNT ( CASE WHEN status = 1 THEN 1 END )     AS complete
     ,       COUNT ( CASE WHEN status = 2 THEN 1 END )      AS incomplete
     FROM        pt_status
     WHERE        status IN (1,2)
     GROUP BY  patient_id
SELECT       patient_id, complete, incomplete
,                                complete / (complete + incomplete)     AS Ratio
,       RANK () OVER ( ORDER BY  complete / (complete + incomplete) ) - 1
                                                    AS decile
FROM      got_complete
ORDER BY  complete
,            incomplete     DESC
;RANK numbers rows 1, 2, 3, ... If I understand your requirements, you want the numbering to start with 0 (meaning "there are 0 other patients with a lower ratio"), so that's why I subtracted 1.
Depending how you want to handle ties, you may need to add some tie-breaker expressions to the analytic ORDER BY clause, and/or use ROW_NUMBER instead of RANK.
Here, I used the calculation "complete / (complete + incomplete)" in the ratio column, then repeated it in the decile column.
I you want, you can add another sub-query, called got_ratio, which would add the ratio column, but do nothing about the decile column. Then, in the main query, you could use RANK as shown above (or, as you suggested, a scalar sub-query referencing got_ratio) to get the number of other pateient_ids with lower ratios.
On the other hand, you could do this whole job without any sub-queries, using AVG as I did earlier, and then repeating that same AVG expression in the ORDER BY clause for RANK. Aggregate functions are computed before analytic fucntions, so the analytic RANK can referenece the aggregate AVG.
I hope this answers your question.
If not, post a little sample data (CREATE TABLE and INSERT statements, relevant columns only), and also post the results you want from that data. Include examples of decile ties (2 or more patients with the same ratio).
Explain, using specific examples, how you get those results from that data.