Comment classer les résultats de deux requêtes SQL?

Question

SELECT * FROM(
     (SELECT 
      count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
      TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
     FROM 
      ALRT_PLATFORM_ALRT_HSTRY 
     where 
      APPL_CD like 'EBP' and 
      ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
      group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 
     ) b, 

     (SELECT 
      count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
      TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
     FROM 
      ALRT_PLATFORM_ALRT 
     where 
      APPL_CD like 'EBP' and 
      ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
      group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 
     ) f 
     ) 

cette requête retourne les données dans le format suivant:

NOOFUSERS    ACCDATE NOOFUSERS    ACCDATE  
---------------------- ---------- ---------------------- ---------- 
2      2009-12-21 1      2010-03-01 
2      2009-12-21 2      2010-03-02 
2      2009-12-21 1      2010-03-03 
1      2009-12-23 1      2010-03-01 

est-il possible au club le reuslt de deux tables: Am attendant que les données soient dans ce format:

NOOFUSERS    ACCDATE  
---------------------- ---------- 
2      2009-12-21 
1      2009-12-23 
1      2010-01-02 
1      2010-01-04 //till here its table one data 
1      2010-03-01 //from here its table TWO data 
2      2010-03-02 
1      2010-03-03 

Merci :)

Answer 1

L'utilisation de UNION ALL à la place des colonnes sous-sélectionnées ajoutera les instructions UNION ALL suivantes à la sélection d'origine. Méfiez-vous, vous avez besoin de la même quantité de colonnes dans chaque instruction select. UNION ALL s'assure que le résultat combiné n'est pas trié/mêlé (la première requête est renvoyée avant la seconde). Une UNION régulière peut/va mélanger/mélanger/trier le résultat.

SELECT 
      count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
      TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
     FROM 
      ALRT_PLATFORM_ALRT_HSTRY 
     where 
      APPL_CD like 'EBP' and 
      ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
      group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 


UNION ALL 

     SELECT 
      count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
      TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
     FROM 
      ALRT_PLATFORM_ALRT 
     where 
      APPL_CD like 'EBP' and 
      ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
      group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 

Answer 2

Peut-être que vous devez utiliser l'opérateur UNION? http://dev.mysql.com/doc/refman/5.5/en/union.html

SELECT count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
      TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
     FROM 
      ALRT_PLATFORM_ALRT_HSTRY 
     where 
      APPL_CD like 'EBP' and 
      ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
      group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 

UNION  
    SELECT count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
      TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
     FROM 
      ALRT_PLATFORM_ALRT 
     where 
      APPL_CD like 'EBP' and 
      ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
      group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 

Answer 3

essayer les déclarations syndicales

(select1) union (select2)

SELECT 
     count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
     TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
    FROM 
     ALRT_PLATFORM_ALRT_HSTRY 
    where 
     APPL_CD like 'EBP' and 
     ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
     group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 

UNION 

    SELECT 
     count(DISTINCT RECEPIENT_ID) as NoOfUsers, 
     TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
    FROM 
     ALRT_PLATFORM_ALRT 
    where 
     APPL_CD like 'EBP' and 
     ALRT_RSPNS_FROM_CLIENT_ID like 'BB' 
     group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') 

Answer 4

Vous pouvez combiner les deux résultats à l'aide UNION

Exemple:

select col1, col2 
from table1 
where... 

UNION 

select col1, col2 
from table2 
where... 

Answer 5

Comme je le vois, vous n'avez pas des conditions de jointure entre les deux ensembles (b et f). Donc, utilisez UNION ou UNION ALL. Refer here.

Answer 6

Si vous avez besoin d'interroger ces tables assez souvent ensemble, vous pouvez envisager d'utiliser une vue.

create view alrt_platform_alrt_all as 
    select * from alrt_platform_alrt 
union all 
    select * from alrt_platform_alrt_hstry; 

Cela vous permettra d'interroger les données comme ceci:

select 
    count(distinct recepient_id), 
    TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD') as accDate 
from alrt_platform_alrt_all 
where appl_cd like 'EBP' 
    and alrt_rspns_from_client_id like 'BB' 
group by TO_CHAR(ACTN_TAKE_DATA_TM,'YYYY-MM-DD');