sharing/swapping items between lists

[Aaron Brady]

On Apr 14, 12:37 pm, Ross <ross.j... at gmail.com> wrote:
> On Apr 14, 10:34 am, Ross <ross.j... at gmail.com> wrote:
>
>
>
> > On Apr 14, 5:57 am, a... at pythoncraft.com (Aahz) wrote:
>
> > > In article <f64c9de2-3285-4f74-adb8-2111c78b7... at 37g2000yqp.googlegroups.com>,
>
> > > Ross  <ross.j... at gmail.com> wrote:
> > > >On Apr 13, 9:08=A0am, a... at pythoncraft.com (Aahz) wrote:
> > > >> In article <c569228f-f391-4317-83a2-08621c601... at r8g2000yql.googlegroups.=
> > > >com>,
> > > >> Ross =A0<ross.j... at gmail.com> wrote:
>
> > > >>>I'm sorry...my example was probably a bad one. A better example of
> > > >>>output I would like would be something like [[1,2],[3,4],[5,6]] and
> > > >>>then for the leftovers list [7,8,9,10 etc]. What I'm trying to do is
> > > >>>produce some sort of round robin algorithm for tennis that is
> > > >>>constrained by the number of courts available each week. So if there
> > > >>>are only 3 courts available for a singles league and 10 people have
> > > >>>signed up, 4 players will have a bye each week. I want my algorithm to
> > > >>>produce unique matchups each week and also give each player the same
> > > >>>angle?
>
> > > >> How about Googling for "round robin algorithm python"? ;-)
>
> > > >I have the basic algorithm and it works fine...I'm just having trouble
> > > >adding another parameter to it that allows for court constraints and
> > > >bye weeks.
>
> > > You'll need to give us more information, then.  Why don't you start with
> > > the core algorithm you're using?
> > > --
> > > Aahz (a... at pythoncraft.com)           <*>        http://www.pythoncraft.com/
>
> > > Why is this newsgroup different from all other newsgroups?
>
> > Here's the core algorithm I'm using:
>
> > >>> def round_robin(teams,rounds):
>
> >         if len(teams)%2:
> >                 teams.append(None)
> >         mid = len(teams) //2
> >         for i in range(rounds):
> >                 yield zip(teams[:mid], teams[mid:])
> >                 teams = teams[0:1] + teams[mid:mid+1] + teams[1:mid-1]+teams[mid
> > +1:]+teams[mid-1:mid]
>
> > >>> if __name__== '__main__':
>
> >         rounds = 15
> >         teams = range(16)
> >         for round in round_robin(teams,rounds):
> >                 print round
>
> fyi rounds=15 and teams =range(16) was just test code I was playing
> around with...they could theoretically be anything.

Here is an idea.  Create a list of all possible pairs, using
itertools.combinations.  You'll notice everyone gets equal play time
and equal time against each other on a pair-by-pair basis.  Then, call
random.shuffle until one player isn't playing on two courts in one
day.

It has the disadvantage that you might end up with player A playing
lots early on and rarely at the end, and B rarely early on and lots at
the end.  Perhaps you could generate a few to several correct
solutions, then choose the most evenly distributed.  Does this make
sense?