Re: columns for count histograms of values
2008-05-01 01:32:51 GMT
Just curious, how are you planning to display the histogram?
Are you allowed to use application code (C/C++/Perl, etc..) to generate the histogram? Personally, SQL is great for showing the data but not good for making graphs with data you can show.
Greetings -- I have a table of the kind
-- where value can take any value in the range 1 to 5. Now I want to have a statistical table Stats of the form
-- how can I create it in one pass over Ratings? I can use min(), max(), avg() for
insert into stats values (id,select min(rating), max(rating), avg(rating), ...) from ratings
-- but what to do for r1,..,r5, short of subselects (select count(rating) from ratings where stats.id=ratings.id) for each, which is an overkill?
Also, if a table Stats already exists with some more columns, and we need to do an update, not insert, for the above, how would that work --
update stats set min=min(ratings), ... from ratings where stats.id=ratings.id -- how do we do the histogram in this case, where the id is fixed explicitly?
Sent via pgsql-sql mailing list (pgsql-sql <at> postgresql.org)
To make changes to your subscription: