CrossJoin with InStr function won't bring in value from column
The code below runs fine and brings in the correct Provider Lookup; however, if I change a."Provider Lookup" to a."Provider", it does not bring back the Provider but plugs in the "N". I can't figure out what's happening... I have confirmed that the HCPC in the WFMGMT_PRD.map_auto_provider table and zip code match to the same instance within the wfmgmt_prd.open_report_hourly
select o."zip code",
MAX(CASE WHEN INSTR(o.'provider hcpc/revenue code', a.HCPC) >= 1 and cast(cast(cast(o.'zip code' as float) as int) as varchar)= a.zip THEN a."Provider Lookup" ELSE 'N' END) "Auto-Provider"
from wfmgmt_prd.open_report_hourly o
CROSS JOIN WFMGMT_PRD.map_auto_provider a
where "Report Interval" >= (select max("Report Interval") from wfmgmt_prd.open_report_hourly)
and o."queue type" = 'Provider Staffing'
and o."intake id"='10969508'
group by o."zip code",o."provider hcpc/revenue code"
Leave a Comment
Can't find what you're looking for? Search the Vertica Documentation, Knowledge Base, or Blog for more information.
Do you mean you hit the issue if you changed a."Provider Lookup" in CASE clause to a."Provider"? If possible, please provide the simple DDLs and data for the reproducer.
My trial worked well. But I think this may be different from yours.
I've attached an Excel document showing the 2 tables queried separately (2 separate tabs) - then the crossjoin that will not bring in the provider name (3rd tab).
Thank you for providing this information.
I tried to reproduce your issue but it failed.
Which Vertica version are you using? Can you try the attached scenario and see if the issue happens?
Vertica Analytic Database v10.1.1-1
We all got together to try to figure this out and we think the issue is Vertica is not bringing back all the rows to match as it must have a limit.
Can you share the actual DDLs for these tables and the data to reproduce the issue in our environment? If yes, please send an email to me. If the data is huge, I can prepare the temporary sftp site for you. As for the data, any data you can see this issue is OK for me.
Your environment is 10.1.1-1. How many nodes does your cluster have?