Re: joins: avoiding recurrences

Hello Daniel Morgan,

I have a very temporary limited job, providing a web interface to a limited selection of tables (only) to our spreaded working group.

If you write me, that (natural joins)
> They are the invention of a marketing department and have no place in
> any serious application.

I couldn't win anything from that. I know how they work, I can see the results... Shall I better use <table join table using(column)>? I know the data well and I - , as I wrote, in lack of much experience, exept for a 3-days-oracle sql course - use natural joins preparing the joint views that way that I have two columns on which it can be performed. I'm not a certified I have to break in, that's live.

I really thank you for your constructive hint on pivot tables for my last problem. So I solved it.
But what I couldn't manage until now is how to merge for instance rows belonging to the same sample (like the first two rows):

well we_id sample_id sample_date O_18 O_18_err H_2 H_2_err

well_1  98976   010189   1989-01-01     5,5       0,2		
well_1  89988   010189   1989-01-01                    22,3      0,4
well_4  26808   013268   2003-01-10     4,3       0,2		

In general I don't expect someone coding my problem, but hints from one who is doing this maybe 10 times a week. So if you ask for posting my code and then answering natural joints are not smart..., then it is more effectively to leave a question open. Be sure I meant that not to blame but to be constructive.

Torsten Received on Mon Feb 23 2004 - 18:31:04 CST