On Wed, 15 Feb 2012 19:20:21 +0100, Franck Ditter wrote: > What is the cost of calling primes(n) below ? I'm mainly interested in > knowing if the call to append is O(1)
Your primes() function appears to be a variation on trial division, which is asymptotically O(n*sqrt(n)/(log n)**2). Regardless of the exact Big Oh behaviour, it is going to be SLOW for large values of n. The behaviour of append is the least of your concerns. -- Steven -- http://mail.python.org/mailman/listinfo/python-list
