I sent this to John directly. Posting to the list. -------- Forwarded Message -------- From: Victor Bouffier <[EMAIL PROTECTED]> To: John Fouhy <[EMAIL PROTECTED]> Subject: Re: [Tutor] Extending a list within a list comprehension Date: Tue, 11 Apr 2006 18:03:41 -0500
On Wed, 2006-04-12 at 10:29 +1200, John Fouhy wrote: > On 12/04/06, Victor Bouffier <[EMAIL PROTECTED]> wrote: > > elements = [ > > (codigo, [ cant, importe, porc]), > > (codigo, [ cant, importe, porc]), > > ... > > ] > > > > And I want to sort descending on 'importe', which is x[1][1] for x in > > elements. > > In python 2.4, you could achieve this by saying: > > elements.sort(key=lambda x: x[1][1]) > > In earlier versions of python, you can do: > > elements.sort(lambda x, y: cmp(x[1][1], y[1][1])) > > (but this is less efficient than key= in 2.4) > > There's also the decorate-sort-undecorate idiom: > > dec = [(x[1][1], x) for x in elements] > dec.sort() > elements = [x[1] for x in dec] Hi John, Thanks for your help. This is very nice and I will definitely use it when sorting a fixed list. However, I forgot to point out I am including an extra element: item description from a dict, where its key=codigo (x[0]). This is why I write the whole list as a single element. Still I could follow your suggestion and do: dec = [(x[1][1], x, description[x[0]]) for x in elements] dec.sort() elements = [x[1], x[2] for x in dec] It is still better, but not quite there. Any feedback will be gladly taken. Victor _______________________________________________ Tutor maillist - Tutor@python.org http://mail.python.org/mailman/listinfo/tutor
