On Friday 2004-08-06 12:42, Michael Hudson wrote:
class C(random.choice([dict, list])): pass
is my favourite example of this :-)
Where is THAT monstrocity used!?
Well, it's not, I hope :-)
Hmmm... maybe if you had some algorithm that was more efficient using either a list or a dict, depending on the input data, but in a way that makes it difficult to tell ahead of time which to use, and you want to improve the average behaviour...
... then you'd do better to find which out of (dict,list) does better on average, and always use that one. Sorry. :-)
By running the above code on a wide range of inputs? :-)
<boring>No, you'd get more reliable results for a given expenditure of effort by running half as many input datasets with class C(dict) and class C(list).</boring> -- g