On Wed, 22 Dec 2004 21:42:16 GMT, Steven Bethard <[EMAIL PROTECTED]> wrote: > Ishwor wrote: > > i am trying to remove an item 'e' from the list l > > I thought it might be helpful to code some of the alternatives you've > been given and look at the timings to put things into perspective. The > code: > > -------------------- remove.py -------------------- > def remove_lc(x, lst): > lst[:] = [item for item in lst if item != x] > > def remove_list(x, lst): > result = [] > for item in lst: > if item != x: > result.append(item) > lst[:] = result > > def remove_filter(x, lst): > lst[:] = filter(lambda item: item != x, lst) > > def remove_xrange(x, lst): > for i in xrange(len(lst)-1, -1, -1): > if lst[i] == x: > del lst[i] > > def remove_remove(x, lst): > while x in lst: > lst.remove(x) > > def remove_try(x, lst): > try: > while True: > lst.remove(x) > except ValueError: > pass > > def remove_ishwor(x, lst): > for item in lst[:]: > if item == x: > lst.remove(x); > -------------------------------------------------- > > First, some timings when only 1 out of every 1000 elements needs to be > removed. Timings were taken with Python 2.4 on a 2.26 Ghz Windows box > using: > $ python -m timeit -s "import remove; lst = [x % 1000 for x in > xrange(10000)]" "remove.remove_<name>(500, lst)" > > remove_remove: 516 usec per loop > remove_try: 604 usec per loop > remove_ishwor: 1.61 msec per loop > remove_xrange: 2.29 msec per loop > remove_lc: 2.37 msec per loop > remove_list: 5.3 msec per loop > remove_filter: 5.65 msec per loop > > Now, some timings when 1 out of every 10 elements needs to be removed. > Timings were taken using: > $ python -m timeit -s "import remove; lst = [x % 10 for x in > xrange(10000)]" "remove.remove_<name>(5, lst)" > > remove_lc: 2.03 msec per loop > remove_xrange: 2.08 msec per loop > remove_list: 4.72 msec per loop > remove_filter: 5.17 msec per loop > remove_try: 30.7 msec per loop > remove_ishwor: 31.5 msec per loop > remove_remove: 60.2 msec per loop > > The moral of the story here is that, if the items to be removed only > make up a very small percentage of the list, an approach like > remove_remove or remove_try might be okay. On the other hand, if you > expect the items to be removed will make up even a moderate percentage > of the list (e.g. 10%), then remove_lc or remove_xrange is a vastly > better alternative. > > Steve > -- > http://mail.python.org/mailman/listinfo/python-list > umm... looks good.. i am running these task sets on my box as well slightly slower than yours 2.4Ghz... thanx Steve. :-) now i'll sit back and dwell a bit deeper into it now.
-- cheers, Ishwor Gurung -- http://mail.python.org/mailman/listinfo/python-list
