+- Python Forum (https://python-forum.io)
+-- Forum: Python Coding (https://python-forum.io/forum-7.html)
+--- Forum: Data Science (https://python-forum.io/forum-44.html)
+--- Thread: Timeit module (/thread-7863.html)
Timeit module - Miraclefruit - Jan-28-2018
I'm trying to see the runtime of s.intersection(t) that takes two sets s and t and returns a new set with all the elements that occur in both s and t.
So, for example, i want to try running |s| = 1000 and |t| = 1000
I tried
import timeit
for x in range(1000):
t = timeit.Timer("s.intersection(t)", "s = set(1000)", "t = set(1000)")
t.timeit()I keep getting errors no matter what variation I try of Timer(). Any ideas?
RE: Timeit module - egslava - Jan-28-2018
s = set(1000)This code is not runnable.
RE: Timeit module - Miraclefruit - Jan-28-2018
(Jan-28-2018, 02:16 AM)egslava Wrote:s = set(1000)This code is not runnable.
how do i fix this? i really can't find any good examples of timeit module using intersection
RE: Timeit module - egslava - Jan-28-2018
%%timeit
import numpy as np
NUM_ELEMS = 10
SCALE = 1000
s1 = set( (np.random.rand(NUM_ELEMS) * SCALE).astype('int') )
s2 = set( (np.random.rand(NUM_ELEMS) * SCALE).astype('int') )
s1.intersection(s2)Quote:The slowest run took 4.28 times longer than the fastest. This could mean that an intermediate result is being cached.
10000 loops, best of 3: 20.8 µs per loop
This code creates two arrays of 10 random items, each is from range [0..1000), then make them sets (so the set size can be less than 1000, since duplicated elements are removed), then, it calculates their intersection.
It's all under %%timeit cell magic, so
timeit will give you the "benchmark" result :)
RE: Timeit module - Miraclefruit - Jan-28-2018
(Jan-28-2018, 02:25 AM)egslava Wrote: %%timeit import numpy as np NUM_ELEMS = 10 SCALE = 1000 s1 = set( (np.random.rand(NUM_ELEMS) * SCALE).astype('int') ) s2 = set( (np.random.rand(NUM_ELEMS) * SCALE).astype('int') ) s1.intersection(s2) The slowest run took 4.28 times longer than the fastest. This could mean that an intermediate result is being cached. 10000 loops, best of 3: 20.8 µs per loop This code creates two arrays of 1000 random elements, then make them sets (so the set size can be less than 1000, since duplicated elements are removed), then, it calculates their intersection.what if i want to do s = 1000 and t = 10,000
RE: Timeit module - snippsat - Jan-28-2018
One of the simplest way beside %%timeit cell magic,is just to put all in string.
import timeit
np_test = '''\
import numpy as np
NUM_ELEMS = 10
SCALE = 1000
s1 = set( (np.random.rand(NUM_ELEMS) * SCALE).astype('int') )
s2 = set( (np.random.rand(NUM_ELEMS) * SCALE).astype('int') )
s1.intersection(s2)
'''
print(timeit.Timer(stmt=np_test).timeit(number=100000))Output:3.0972112079055183RE: Timeit module - Miraclefruit - Jan-28-2018
How does that take the intersection into consideration?
RE: Timeit module - egslava - Jan-28-2018
To be honest, I can't really understand your question, what is 't' and what is 's'?
You mean, that one set should have 1000 random elements and the other set 10,000? This way then:
In [1]: %%timeit
...:
...: import numpy as np
...: SCALE = 1000
...: s1 = set( (np.random.rand(1000) * SCALE).astype('int') )
...: s2 = set( (np.random.rand(10000) * SCALE).astype('int') )
...: s1.intersection(s2)
...:
The slowest run took 748.40 times longer than the fastest. This could mean that
an intermediate result is being cached.
1 loop, best of 3: 1.28 ms per loop
RE: Timeit module - snippsat - Jan-28-2018
(Jan-28-2018, 03:43 AM)Miraclefruit Wrote: How does that take the intersection into consideration?What code dos do not matter at all for what i show,the point was to show an easy way to run timeit.
Tuple are faster than list,and boost number executions to 100000000
import timeit tuple_test = '''\ t = (1,2,3,4,5,6,7,8,9,10) ''' list_test = '''\ t = [1,2,3,4,5,6,7,8,9,10] ''' print(timeit.Timer(stmt=tuple_test).timeit(number=100000000))
RE: Timeit module - Miraclefruit - Jan-28-2018
Got it! Thanks.
Yeah they're just 2 different sets