I've read some on table partitioning and using nested select statements
with group by, but have not found the syntax to produce the needed results.
From a table I extract row counts grouped by three columns:
select stream, sampdate, func_feed_grp, count(*) from benthos group
by stream, sampdate, func_feed_grp order by stream, sampdate,
func_feed_group;
And I want to include the proportion of each count based on the total rows
for each stream and sampdate. The totals are obtained with this statement:
select stream, sampdate, count(*) as tot_cnt from benthos group by stream,
sampdate order by stream, sampdate;
What I do not know how to do is combine the two so the resulting table
contains the columns stream, sampdate, count, proportion. I want to learn
how to build the sub-select to get this result. Joe Celko's 'SQL for
Smarties, 4th Ed.' has a close example in the chapter on table partitioning,
but I could not apply that model to my table.
TIA,
Rich
