def single_algo(gen): dd = np.zeros((1 + repeats, len(b_params))) dd[0, :] = b_params #to fill algo = kadane probeIndex = 0 too_deep = False for probe in b_params: if too_deep: break print 'Probe size ', probe, '(Index ', probeIndex, ')...' for row in range(1, repeats + 1): print 'Repeat ', row, '...' data = gen(100000, max_val=probe) try: perf = measure_exe_time(algo, data) dd[row, probeIndex] = perf print perf except RuntimeError: dd[row, probeIndex] = 0 too_deep = True probeIndex += 1 return dd
def f(n, b): results = [] for i in range(1): data = generate_random_sequence(n, max_val = b) perf = measure_exe_time(counting_sort, data) results.append(perf) return mean(results)
def single_algo(gen):
dd = np.zeros((1 + repeats, len(probe_sizes)))
dd[0, :] = probe_sizes
# to fill
algo = shell_sort
probeIndex = 0
too_deep = False
for probe in probe_sizes:
if too_deep:
break
print "Probe size ", probe, "(Index ", probeIndex, ")..."
for row in range(1, repeats + 1):
print "Repeat ", row, "..."
data = gen(probe)
try:
perf = measure_exe_time(algo, data)
dd[row, probeIndex] = perf
print perf
except RuntimeError:
dd[row, probeIndex] = 0
too_deep = True
probeIndex += 1
return dd
def bench():
probes = range(1, 21)
repeats = 1
rec = np.zeros((len(probes), 3))
ite = np.zeros((len(probes), 3))
ri = []
row = 0
for probe in probes:
print probe
rr = []
for i in range(repeats):
rtime = measure_exe_time(generate_gray_codes, 2 ** probe)
ri.append(rtime)
#ri.append(measure_exe_time(generate_gray_codes, probe))
rr.append(measure_exe_time(wdi, probe))
rmean = np.mean(rr)
rmax = np.max(rr)
rmin = np.min(rr)
rec[row, :] = [rmean, rmax, rmin]
imean = np.mean(ri)
imax = np.max(ri)
imin = np.min(ri)
ite[row, :] = [imean, imax, imin]
row += 1
p1, = plt.plot(probes, ite[:, 0])
p2, = plt.plot(probes, rec[:, 0])
plt.legend([p1, p2], ['Iterative', 'Recursive'], loc=2)
plt.xlabel('Number of bits in code')
plt.ylabel('Execution time [s]')
plt.savefig('gray.png')
plt.show()
def single_algo(gen): dd = np.zeros((len(b_param), len(probe_sizes))) #dd[0, :] = probe_sizes #dd[:, 0] = b_param #to fill algo = counting_sort nc = 0 bc = 0 for n in probe_sizes: bc = 0 for b in b_param: print 'n=', n, ' b=', b r = [] for xd in range(repeats): data = generate_random_sequence(n, max_val=b) perf = measure_exe_time(algo,data) r.append(perf) dd[bc, nc] = np.mean(r) bc += 1 nc += 1 '''probeIndex = 0 too_deep = False for probe in probe_sizes: if too_deep: break print 'Probe size ', probe, '(Index ', probeIndex, ')...' for row in range(1, repeats + 1): print 'Repeat ', row, '...' data = gen(probe) try: perf = measure_exe_time(algo, data) dd[row, probeIndex] = perf print perf except RuntimeError: dd[row, probeIndex] = 0 too_deep = True probeIndex += 1''' return dd
return int(raw_input()) if __name__ == '__main__': algo = menu_algo() size = menu_size() generator = menu_generator() algos = [counting_sort, QuickSort, QuickSortIterative, heap_sort, shell_sort] generators = [generate_ascending_sequence, generate_descending_sequence, generate_v_sequence, generate_random_sequence] seq = generators[generator](size) if size < 11: print "Sekwencja wejsciowa: ", seq perf = measure_exe_time(algos[algo], seq) if size < 11: print "Sekwencja wyjsciowa: ", seq print "Czas wykonania: ", perf print is_sorted_desc(seq) raw_input()
