try this

select id, avg(sal) from @temp

where sal <> 0

group by id

having count(reportingperiod) = 2

declare variable to store reportingperiod. in above example i have used constant 2

Statement like you shown is not showing any cascading join. By cascading join i means A --> B --> C. If cascading join is used it will be processed by...

Jeff,

The method i can straight away suggest is to use SSIS.

1. Create in-memory single column table with each csv row as table row

2. find COMMA's between "xxxx,yyyy"

3. replace with...

cimbom_kid

in that case you can add another update statment which will change group_ID based on creteria previously inserted group header

CREATE TABLE #TMP_LOADER_2 (

group_id int IDENTITY (1, 1)

,linenum int)

insert into #TMP_LOADER_2

select...

Try this one:

CREATE TABLE #TMP_LOADER_2 (

group_id int IDENTITY (1, 1)

,linenum int)

insert into #TMP_LOADER_2

select linenum from #TMP_LOADER where fileline = '-----'

UPDATE t1

SET t1.group_id = t2.group_id

FROM #TMP_LOADER t1 inner join #TMP_LOADER_2...

i had same issue where we use to import CSV files. Best option i would recommend is first clean up your CSV by removing columns with internal COMMA's

Good luck