I am trying to execute a query in hive and getting the error. I check over and over again but I cannot see any problem.
select
a.phone_no,
a.app_name
from
(select * from (select app_name,phone_no from lc_app_flag) ) a
inner join
(select * from (select phone_no,city_id_day,city_id_night,lat_day,lng_day,lat_night,lat_night from TW_FEATS_FIN_LCEXT01 where month_id='202205' ) b where city_id_day='440100' or city_id_night='440100') c
on a.phone_no=c.phone_no
You missed aliasing first innermost subquery. Correct SQL below -
select
a.phone_no,
a.app_name
from
(select * from (select app_name,phone_no from lc_app_flag) subq) a --inner subquery as subq
inner join
(select * from (select phone_no,city_id_day,city_id_night,lat_day,lng_day,lat_night,lat_night from TW_FEATS_FIN_LCEXT01 where month_id='202205' ) b where city_id_day='440100' or city_id_night='440100') c
on a.phone_no=c.phone_no
Related
I'm getting a syntax error at Left Join. So in trying to combine the two, i used the left join and the brackets. I'm not sure where the problem is:
SELECT DISTINCT a.order_id
FROM fact.outbound AS a
ORDER BY Rand()
LIMIT 5
LEFT JOIN (
SELECT
outbound.marketplace_name,
outbound.product_type,
outbound.mpid,
outbound.order_id,
outbound.sku,
pbdh.mpid,
pbdh.product_type,
pbdh.validated_exp_reach,
pbdh.ultimate_sales_rank_de,
pbdh.ultimate_sales_rank_fr,
(
pbdh.very_good_stock_count + good_stock_count + new_Stock_count
) as total_stock
FROM
fact.outbound AS outbound
LEFT JOIN reporting_layer.pricing_bi_data_historisation AS pbdh ON outbound.mpid = pbdh.mpid
AND trunc(outbound.ordered_date) = trunc(pbdh.importdate)
WHERE
outbound.ordered_date > '2022-01-01'
AND pbdh.importdate > '2022-01-01'
LIMIT
5
) AS b ON a.orderid = b.order_id
Error:
You have an error in your SQL syntax; it seems the error is around: 'LEFT JOIN ( SELECT outbound.marketplace_name, outbound.product_t' at line 9
What could be the reason?
Place the first limit logic into a separate subquery, and then join the two subqueries:
SELECT DISTINCT a.order_id
FROM
(
SELECT order_id
FROM fact.outbound
ORDER BY Rand()
LIMIT 5
) a
LEFT JOIN
(
SELECT
outbound.marketplace_name,
outbound.product_type,
outbound.mpid,
outbound.order_id,
outbound.sku,
pbdh.mpid,
pbdh.product_type,
pbdh.validated_exp_reach,
pbdh.ultimate_sales_rank_de,
pbdh.ultimate_sales_rank_fr,
(pbdh.very_good_stock_count +
good_stock_count + new_Stock_count) AS total_stock
FROM fact.outbound AS outbound
LEFT JOIN reporting_layer.pricing_bi_data_historisation AS pbdh
ON outbound.mpid = pbdh.mpid AND
TRUNC(outbound.ordered_date) = TRUNC(pbdh.importdate)
WHERE outbound.ordered_date > '2022-01-01' AND
pbdh.importdate > '2022-01-01'
-- there should be an ORDER BY clause here...
LIMIT 5
) AS b
ON a.orderid = b.order_id;
Note that the select clause of the b subquery can be reduced to just the order_id, as no values from this subquery are actually selected in the end.
You can skip the LEFT JOIN, since no b columns are selected. (And SELECT DISTINCT makes sure any duplicates are eliminated.)
SELECT DISTINCT order_id
FROM fact.outbound
ORDER BY Rand()
LIMIT 5
I am trying to run my query but I get an error.
This is my query:
if exists (select CODE_ISIN
from cte
where code_ISIN not in (select [STATUT_TITRE], [CODE_ISIN]
from TT_TITRE A
inner join TT_STATUT_TITRE B on A.TITRE_ID = B.TITRE_ID))
begin
select 'ko'
end
begin
select 'ok'
end
Remove [STATUT_TITRE] from sub-query as it will accept only one expression :
select c.CODE_ISIN
from cte c
where code_ISIN not in (select [CODE_ISIN] -- only one expression needed
from TT_TITRE A inner join
TT_STATUT_TITRE B
on A.TITRE_ID = B.TITRE_ID
);
I would suggest to use NOT EXISTS instead :
where not exists (select 1
from TT_TITRE A inner join
TT_STATUT_TITRE B
on A.TITRE_ID=B.TITRE_ID
where CODE_ISIN = c.CODE_ISIN
);
I'm executing this query in SqlDeveloper targeting a database in Oracle 10g:
select P1.FIELD1, P1.COD from
(select P1.FIELD1, P1.COD, row_number()
over (order by P1.FIELD1) r from
SCHEMA1.P2
INNER JOIN SCHEMA1.P1 on SCHEMA1.P1.COD=P2.FIELD2)
where P1.FIELD_11 = 'VALUE1' and r between 6 and 10;
And Oracle returns:
Error SQL: ORA-00904: "P1"."FIELD_11": invalid identifier
I tried:
select SCHEMA1.P1.FIELD1, SCHEMA1.P1.COD from
(select SCHEMA1.P1.FIELD1, SCHEMA1.P1.COD, row_number()
over (order by SCHEMA1.P1.FIELD1) r from
SCHEMA1.P2
INNER JOIN SCHEMA1.P1 on SCHEMA1.P1.COD=SCHEMA1.P2.FIELD2)
where SCHEMA1.P1.FIELD_11 = 'VALUE1' and r between 6 and 10;
And I got the same error:
ORA-00904: "SCHEMA1"."P1"."FIELD_11": invalid identifier
I cannot find examples of select clauses with inner join and pagination for Oracle 10g, so I don't know what I'm doing wrong.
You have messed the query with aliases which makes it confusing.use shorter aliases instead of whole schema.table.column everywhere
SELECT p.field1,
p.cod
FROM (SELECT p1.field1,
p1.cod,
p1.field_11,
row_number()
over (
ORDER BY p1.field1) r
FROM schema1.p2 p2
inner join schema1.p1 p1
ON p1.cod = p2.field2) p
WHERE p.field_11 = 'VALUE1'
AND r BETWEEN 6 AND 10;
SELECT *
FROM
construction AS T2
INNER JOIN
(
SELECT project,MAX(report_date)
FROM construction
GROUP BY project
) AS R
ON T2.project=R.project AND T2.report_date=R.report_date
getting this error. plz help
No column name was specified for column 2 of 'R'
You need to add alias for MAX(report_date):
SELECT *
FROM construction AS T2
INNER JOIN
(
SELECT project,MAX(report_date) AS report_date
FROM construction
GROUP BY project
) AS R
ON T2.project = R.project
AND T2.report_date = R.report_date;
In SQL Server you can use syntax:
SELECT *
FROM construction AS T2
INNER JOIN
(
SELECT project,MAX(report_date)
FROM construction
GROUP BY project
) AS R(project, report_date)
ON T2.project = R.project
AND T2.report_date = R.report_date;
You should specific the MAX(report_date) with an alias report_date.
Because your table R have two columns project,MAX(report_date).
You are getting this error because you have not specified column name for inner query
You have to write your query as
SELECT *
FROM construction
INNER JOIN
(
SELECT project,MAX(report_date)"Max_ReportDate"
FROM construction
GROUP BY project
) Max_construction
ON construction.project = Max_construction .project
AND construction.report_date = Max_construction .Max_ReportDate
I have the below query in my application which was running on DB2:
SELECT COD.POST_CD,CLS.CLASS,COD2.STATUS_CD
FROM DC01.POSTAL_CODES COD
INNER JOIN DC02.STATUS_CODES COD2
ON COD.ORDER=COD2.ORDER
INNER JOIN DC02.VALID_ORDERS ORD
ON ORD.ORDER=COD.ORDER
WHERE
(
( EXISTS (SELECT 1 FROM DC00.PROCESS_ORDER PRD
WHERE PRD.ORDER=COD.ORDER
AND PRD.IDNUM=COD.IDNUM
)
) OR
( EXISTS (SELECT 1 FROM DC00.PENDING_ORDER PND
WHERE PND.ORDER=COD.ORDER
AND PND.IDNUM=COD.IDNUM
)
)
)
AND EXISTS (SELECT 1 FROM DC00.CUSTOM_ORDER CRD
WHERE CRD.ORDER=COD.ORDER
)
;
When we changed to UDB (LUW v9.5) we are getting the below warning:
IWAQ0003W SQL warnings were found
SQLState=01602 Performance of this complex query might be sub-optimal.
Reason code: "3".. SQLCODE=437, SQLSTATE=01602, DRIVER=4.13.111
I know this warning is due to the EXISTS () OR EXISTS statements. But I am not sure any other way I can write this query to replace. If it is AND, I could have made an INNER JOIN, but I am not able to change this condition as it is OR. Can any one suggest better way to replace these EXISTS Statements?
SELECT COD.POST_CD,CLS.CLASS,COD2.STATUS_CD
FROM DC01.POSTAL_CODES COD
INNER JOIN DC02.STATUS_CODES COD2
ON COD.ORDER=COD2.ORDER
INNER JOIN DC02.VALID_ORDERS ORD
ON ORD.ORDER=COD.ORDER
WHERE
(
EXISTS SELECT 1 FROM
(SELECT ORDER,IDNUM FROM DC00.PROCESS_ORDER PRD UNION
SELECT ORDER,IDNUM FROM DC00.PENDING_ORDER PND) PD
WHERE PD.ORDER=COD.ORDER
AND PD.IDNUM=COD.IDNUM
)
AND EXISTS (SELECT 1 FROM DC00.CUSTOM_ORDER CRD
WHERE CRD.ORDER=COD.ORDER
)
;