Hvis du antager, at du vil gruppere dataene, før du genererer nøglen med sekvensen, lyder det, som om du vil have noget lignende
INSERT INTO HISTORICAL_CAR_STATS (
HISTORICAL_CAR_STATS_ID,
YEAR,
MONTH,
MAKE,
MODEL,
REGION,
AVG_MSRP,
CNT)
SELECT MY_SEQ.nextval,
year,
month,
make,
model,
region,
avg_msrp,
cnt
FROM (SELECT '2010' year,
'12' month,
'ALL' make,
'ALL' model,
REGION,
sum(AVG_MSRP*COUNT)/sum(COUNT) avg_msrp,
sum(cnt) cnt
FROM HISTORICAL_CAR_STATS
WHERE YEAR = '2010'
AND MONTH = '12'
AND MAKE != 'ALL'
GROUP BY REGION)