Heavy python list slicing _will_ fuck efficiency, if applicable consider numpy arrays to do the job

slicing_test-numpy_arrays_vs_lists

mpaolino@guacamole:~$ cat bench_slice.py 
import numpy
import timeit

def splitit(arraylist):
    step = 10
    frame = 160
    while len(arraylist) >= frame:
        newlist = arraylist[:frame]
        arraylist = arraylist[step:]

for i in [200, 500, 1000, 10000, 100000]:
    print '-'*80
    print 'N =', i
    nparray = numpy.linspace(0, 1, i)
    pylist = nparray.tolist()
    print "List:", timeit.timeit(stmt='splitit(pylist)', 
                   setup='from __main__ import splitit, nparray, pylist',
                   number=10)
    print "NP array:", timeit.timeit(stmt='splitit(nparray)', 
                       setup='from __main__ import splitit, nparray, pylist',
                       number=10)

mpaolino@guacamole:~$ python bench_slice.py 
--------------------------------------------------------------------------------
N = 200
List: 0.000123023986816
NP array: 7.41481781006e-05
--------------------------------------------------------------------------------
N = 500
List: 0.000946044921875
NP array: 0.000421047210693
--------------------------------------------------------------------------------
N = 1000
List: 0.00359988212585
NP array: 0.000972986221313
--------------------------------------------------------------------------------
N = 10000
List: 0.346795797348
NP array: 0.0114970207214
--------------------------------------------------------------------------------
N = 100000
List: 153.226103067
NP array: 0.133377790451