vsql/Vertica: Select top 5 rows by group

Question

I have this query that it's supposed to take first n rows of grouped data. I used RANK() in combination with OVER PARTITION BY to identify top n rows per group:

SELECT X.USERID, X.ARTID, X.AVGTIMEONPAGE,EDP.ARTDSC,
RANK() OVER (PARTITION BY X.USERID ORDER BY X.AVGTIMEONPAGE DESC) as rank
FROM
(SELECT GANG.userID AS USERID,GANG.avgTimeOnPage AS AVGTIMEONPAGE,   
split_part(GANG.pageTitle,' -',1) as ARTID
FROM GoogleAnalytics.navigazioneG AS GANG
WHERE GANG.pagePath LIKE '%DataSheets%' ) AS X
LEFT JOIN ESPDDS.ESP_DPRODUCT AS EDP
ON EDP.ARTID=X.ARTID AND EDP.SCD_IS_CURRENT=1
AND EDP.COMPANYID=1
WHERE X.ARTID NOT LIKE '%Company%' AND rank in (1,2,3,4,5)

it gives me an error saying that rank column doesn't exist. If I comment the last part of WHERE Clause I can see that column rank is computed correctly.

Thanks

Answer 1

The WHERE clause gets evaluated before the SELECT clause. So at that time rank is unknown. You can use a further subquery to access it:

SELECT *
FROM
(
  SELECT 
    X.USERID, 
    X.ARTID, 
    X.AVGTIMEONPAGE,
    EDP.ARTDSC,
    RANK() OVER (PARTITION BY X.USERID ORDER BY X.AVGTIMEONPAGE DESC) as rank
  FROM
  (
    SELECT 
      GANG.userID AS USERID,
      GANG.avgTimeOnPage AS AVGTIMEONPAGE,   
      split_part(GANG.pageTitle,' -',1) as ARTID
    FROM GoogleAnalytics.navigazioneG AS GANG
    WHERE GANG.pagePath LIKE '%DataSheets%' 
  ) AS X
  LEFT JOIN ESPDDS.ESP_DPRODUCT AS EDP ON EDP.ARTID = X.ARTID
                                      AND EDP.SCD_IS_CURRENT = 1 
                                      AND EDP.COMPANYID = 1
  WHERE X.ARTID NOT LIKE '%Company%' 
) ranked
WHERE rank in (1,2,3,4,5);