Oracle query using DISTINCT
Hello expert,
Using the query copied herewith below yields 4758 rows
Running a similar query with
Select DISITNCT PAYER_ID
FROM CCM.CCM_COMPONENT_PRICING
where last_update_user = 1
yields 202 rows.
The 202 rows is needed but with the other fields
resultant from the longer query.
Tried creating a join with short table against the long table
and tried listing all 202 PAYER_IDs in an 'IN' statement
but both grab all of the PAYER IDs
Is there someway to mimic the return of the DISTINCT
query but return the other fields also?
Thanks
Allen in Dallas
++++++++++++++++LONG QUERY++++++++++++++
Select
COMPONENT_PRICING_ID,
COMPONENT_ID,
PAYER_ID,
VARY_BY_MARKET,
COMPONENT_PRICE,
IS_OVERRIDE,
CREATION_USER,
CREATION_DATE,
LAST_UPDATE_USER,
LAST_UPDATE_DATE
from
CCM.CCM_COMPONENT_PRICING
where last_update_user = 1
Using the query copied herewith below yields 4758 rows
Running a similar query with
Select DISITNCT PAYER_ID
FROM CCM.CCM_COMPONENT_PRICING
where last_update_user = 1
yields 202 rows.
The 202 rows is needed but with the other fields
resultant from the longer query.
Tried creating a join with short table against the long table
and tried listing all 202 PAYER_IDs in an 'IN' statement
but both grab all of the PAYER IDs
Is there someway to mimic the return of the DISTINCT
query but return the other fields also?
Thanks
Allen in Dallas
++++++++++++++++LONG QUERY++++++++++++++
Select
COMPONENT_PRICING_ID,
COMPONENT_ID,
PAYER_ID,
VARY_BY_MARKET,
COMPONENT_PRICE,
IS_OVERRIDE,
CREATION_USER,
CREATION_DATE,
LAST_UPDATE_USER,
LAST_UPDATE_DATE
from
CCM.CCM_COMPONENT_PRICING
where last_update_user = 1
Do more with
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
If you have duplicate payer_ids and some of the other column has different data, which of the rows is valid?
Of you want any of them then this might work:
It basically sorts by the payer_id column and returns the first one it finds.
If you want the payer_id by the last updated data just change the row_number function to:
row_number() over(partition by PAYER_ID order by payer_id, last_update_date desc) rn
Of you want any of them then this might work:
Select
COMPONENT_PRICING_ID,
COMPONENT_ID,
PAYER_ID,
VARY_BY_MARKET,
COMPONENT_PRICE,
IS_OVERRIDE,
CREATION_USER,
CREATION_DATE,
LAST_UPDATE_USER,
LAST_UPDATE_DATE
from
(
Select
COMPONENT_PRICING_ID,
COMPONENT_ID,
PAYER_ID,
VARY_BY_MARKET,
COMPONENT_PRICE,
IS_OVERRIDE,
CREATION_USER,
CREATION_DATE,
LAST_UPDATE_USER,
LAST_UPDATE_DATE,
row_number() over(partition by PAYER_ID order by payer_id) rn
from
CCM.CCM_COMPONENT_PRICING
where last_update_user = 1
)
where rn=1
It basically sorts by the payer_id column and returns the first one it finds.
If you want the payer_id by the last updated data just change the row_number function to:
row_number() over(partition by PAYER_ID order by payer_id, last_update_date desc) rn
Is there someway to mimic the return of the DISTINCTNo. How would the query engine choose which value of the other columns to return?
query but return the other fields also?
If you really don't care which value of COMPONENT_PRICING_ID for example to return, then use a totals query. Use the Group By argument for the "distinct" columns and choose a different aggregation function for all the other columns, First, Last, Min, Max are all possible or you can use Sum, Avg, etc for numeric fields if those functions make sense.
Premium Content
You need an Expert Office subscription to comment.Start Free Trial