IT Answers » GROUP BY statement

SQL Group by Query with aggregate function as well as detail information

idontknowthis — Mon, 08 Oct 2012 11:13:19 +0000

New Answer by TomLiotta

Keep the “first” value using groupby – MYSQL

Iamjeannie — Thu, 14 Oct 2010 19:51:28 +0000

Input :

col1 col2

a      1

b       1

c        2

d       2

I would like to :

Create the new colX to get the first value in col1 group by col2 and save the value in colX

results :

col1 col2   colX

a      1       a

b       1      a

c        2      c

d       2       c

how can do it in mysql

Query performance and tuning

Mingley — Sun, 21 Dec 2008 09:34:55 +0000

Hi,

I wanted know as to what I can do to my query so that it improves in performance.RIght now it it taking 3-4 hrs because of the large no of records in the DB.If i remove the group by and order by clause for the same no of records it is fast but when it does sorting it hangs.But I need to do the sorting a s i am generating a report out of it.Here is my query, note that a lot more conditions are there in the group by and order by claue. I need a quick solution for this.

SELECT A.ACM_AM_ID,B.TCUS_ID,B.CUS_NAME,A.ACM_AC_NUM, A.ACM_ADT_UNIT,C.AAC_USER_ID,F.TFAC_ID,F.FAC_FAC_TYPE,

DECODE(F.FAM_SL_NUM, NULL, ”, F.FAM_SL_NUM), F.FAM_SGD,DECODE(F.FAM_FAM_TYPE, ‘AL’, NVL(F.FAM_NEW_LIMIT, 0) * 10, NVL(F.FAM_NEW_LIMIT, 0)) FAM_NEW_LIMIT ,
FUN_GET_FAC_PURPOSE(F.TFAM_ID, F.FAM_PROD_TYPE), F.FAM_OTH_PURPOSE ,NVL(INTEREST.FPD_SPREAD_RATE, 0,
S.TSIM_ID,
S.SIM_SRTY_TYPE,S.SIM_SRTY_TYPE_OTH,S.SIM_CAT_ASSET,S.SIM_CAT_ASSET_OTH,S.SIM_SGD,W.WIS_ACM_ID, W.WIS_TYPE ,W.WIS_WISSM_FAM_ID PARENT_ID ,W.TFAM_ID CHILD_ID, W.WIS_SGD,
W.WIS_AMT, FAM_LIMIT.AMOUNT ,INTEREST.FPD_INTR_DESC,INTEREST.FPD_VARIANCE

FROM T_ACM_DETAIL A,T_FAM F,T_FAM_WIS_SM W, T_CUS_EMP CE,T_CUS_CORCUS_INFO CU,T_CUS_CUSTOMER B,T_SAM_SECURITY S,

V_LIMIT_BY_ACM FAM_LIMIT, V_FAM_INTEREST INTEREST

WHERE (A.ACM_AM_ID= B.ACM_AM_ID) AND
(FAM_LIMIT.ACM_ADT_ID= A.ACM_AM_ID) AND
(A.ACM_AM_ID= F.ACM_AM_ID) AND (
CE.TCUS_ID(+) = B.TCUS_ID) AND
(CU.TCUS_ID(+) = B.TCUS_ID) AND
(B.ACM_AM_ID= F.ACM_AM_ID) AND
(W.TFAM_ID (+) = F.TFAM_ID ) AND
(INTEREST.TFAM_ID (+) = F.TFAM_ID )AND
(S.ACM_AM_IDIS NOT NULL) AND
(A.ACM_AM_ID= S.ACM_AM_ID(+)) AND (B.ACM_AM_ID= S.ACM_AM_ID)

(A.ACM_AM_ID= B.ACM_AM_ID) AND
(FAM_LIMIT.ACM_AM_ID= A.ACM_AM_ID) AND
(A.ACM_AM_ID= F.ACM_AM_ID) AND
(CE.TCUS_ID(+) = B.TCUS_ID) AND
(CU.TCUS_ID(+) = B.TCUS_ID) AND
(B.ACM_AM_ID= F.ACM_AM_ID) AND
(W.TFAM_ID (+) = F.TFAM_ID ) AND
(INTEREST.TFAM_ID (+) = F.TFAM_ID ) AND
(S.TSIM_ID IS NULL) AND (S.ACM_AM_ID IS NULL) AND (A.ACM_AM_ID= S.ACM_AM_ID(+))

GROUP BY A.ACM_AM_UNIT,A.ACM_AM_ID,B.TCUS_ID
ORDER BY A.ACM_AM_UNIT,A.ACM_AM_ID,B.TCUS_ID

Count several columns Group by accounts Greater than value Beginner

Ben Pinto — Thu, 17 Apr 2008 22:58:55 +0000

I need to be able to count how many values are greater than 1 for each StuNum. I have only been able to count how many values are greater than 1 for each StuNum for d1 or for d2 or for d3, but not for d1, d2 and d3. How can I have it count multiple columns.

Table

StuNum – SchNum – Tchnum – d1 – d2 – d3 – d4 – d5 – d6
999999 – 101 – 4545 – 1 – 0 – 2 – 0 – 1 – 3
999999 – 202 – 3636 – 0 – 1 – 0 – 2 – 1 – 0
444444 – 202 – 3636 – 0 – 1 – 0 – 0 – 0 – 1
555555 – 303 – 7878 – 0 – 2 – 3 – 0 – 0 – 1

SELECT COUNT(d1) AS Doubles, StuNum, SchNum, TchNum
FROM table
WHERE (d1 > 1)
GROUP BY StuNum, SchoolNum, TeachNum

Thanks,