I am looking for simple solution which can help me use full power of my PC to process my data. I think, dividing task onto different core would help in reducing in processing time, but I donot know how to do it, I have searched on stackoverflow for the similar problem but not any solution could resolve my problem. I am processing data of around length: 3000 and since I am using nested for loop to find the number of similar(in +- 0.5 range) elements in the list, it will run 3000x3000 times which takes around 2 minutes and I want to reduce the time taken.
repeat= []
values = []
for i in completeList:
count = 0
for j in completeList:
if isfloat(i) and isfloat(j):
if float(i)-0.5 <= float(j) <= float(i)+0.5:
count = count + 1
repeat.append(count)
values.append(i)
Any help would be appreciated.
with regards, Manish
Since you still did not post the actual code for isfloat or show what the elements of completeList look like, the best I can do is conjecture on what they might be. It makes a difference because as I mentioned, the more CPU required to execute isfloat and float to convert the elements of completeList, the greater the gains to be had by using multiprocessing.
For CASE 1 I am assuming that completeList is composed of strings and that isfloat needs to use a regular expression to determine whether the string matches our expected floating point format and that float therefore needs to convert from a string. This would be what I would imagine to be the most CPU-intensive case. For CASE 2 completeList is composed of floats, isfloat just returns True and float does not have to do any real conversion.
My desktop has 8 core processors:
CASE 1
import multiprocessing as mp
import time
import random
import re
from functools import partial
def isfloat(s):
return not re.fullmatch(r'\d*\.\d+', s) is None
def single_process(complete_list):
#repeat = []
values = []
for idx_i, v_i in enumerate(complete_list):
count = 0
for idx_j, v_j in enumerate(complete_list):
if idx_i == idx_j:
continue # don't compare an element with itself
if isfloat(v_i) and isfloat(v_j):
f_i = float(v_i)
if f_i-0.5 <= float(v_j) <= f_i+0.5:
count = count + 1
# repeat will end up being a copy of complete_list
# why are we doing this?
#repeat.append(v_i)
values.append(count) # these are actually counts
return values
def multi_worker(complete_list, index_range):
values = []
for idx_i in index_range:
v_i = complete_list[idx_i]
count = 0
for idx_j, v_j in enumerate(complete_list):
if idx_i == idx_j:
continue # don't compare an element with itself
if isfloat(v_i) and isfloat(v_j):
f_i = float(v_i)
if f_i-0.5 <= float(v_j) <= f_i+0.5:
count = count + 1
values.append(count) # these are actually counts
return values
def multi_process(complete_list):
def split(a, n):
k, m = divmod(len(a), n)
return (a[i * k + min(i, m):(i + 1) * k + min(i + 1, m)] for i in range(n))
n = len(complete_list)
POOL_SIZE = mp.cpu_count()
range_splits = split(range(0, n), POOL_SIZE)
pool = mp.Pool(POOL_SIZE)
value_lists = pool.map(partial(multi_worker, complete_list), range_splits)
values = []
# join results together:
for value_list in value_lists:
values.extend(value_list)
return values
def main():
# generate 3000 random numbers:
random.seed(0)
complete_list = [str(random.uniform(1.0, 3.0)) for _ in range(3000)]
t = time.time()
values = single_process(complete_list)
print(time.time() - t, values[0:10], values[-10:-1])
t = time.time()
values = multi_process(complete_list)
print(time.time() - t, values[0:10], values[-10:-1])
# required for Windows:
if __name__ == '__main__':
main()
Prints:
27.7540442943573 [1236, 1491, 1464, 1477, 1494, 1472, 1410, 1450, 1502, 1537] [1485, 1513, 1513, 1501, 1283, 1538, 804, 1459, 1457]
7.187546253204346 [1236, 1491, 1464, 1477, 1494, 1472, 1410, 1450, 1502, 1537] [1485, 1513, 1513, 1501, 1283, 1538, 804, 1459, 1457]
CASE 2
import multiprocessing as mp
import time
import random
from functools import partial
def isfloat(s):
return True
def single_process(complete_list):
values = []
for idx_i, v_i in enumerate(complete_list):
count = 0
for idx_j, v_j in enumerate(complete_list):
if idx_i == idx_j:
continue # don't compare an element with itself
if isfloat(v_i) and isfloat(v_j):
f_i = float(v_i)
if f_i-0.5 <= float(v_j) <= f_i+0.5:
count = count + 1
values.append(count) # these are actually counts
return values
def multi_worker(complete_list, index_range):
values = []
for idx_i in index_range:
v_i = complete_list[idx_i]
count = 0
for idx_j, v_j in enumerate(complete_list):
if idx_i == idx_j:
continue # don't compare an element with itself
if isfloat(v_i) and isfloat(v_j):
f_i = float(v_i)
if f_i-0.5 <= float(v_j) <= f_i+0.5:
count = count + 1
values.append(count) # these are actually counts
return values
def multi_process(complete_list):
def split(a, n):
k, m = divmod(len(a), n)
return (a[i * k + min(i, m):(i + 1) * k + min(i + 1, m)] for i in range(n))
n = len(complete_list)
POOL_SIZE = mp.cpu_count()
range_splits = split(range(0, n), POOL_SIZE)
pool = mp.Pool(POOL_SIZE)
value_lists = pool.map(partial(multi_worker, complete_list), range_splits)
values = []
# join results together:
for value_list in value_lists:
values.extend(value_list)
return values
def main():
# generate 3000 random numbers:
random.seed(0)
complete_list = [random.uniform(1.0, 3.0) for _ in range(3000)]
t = time.time()
values = single_process(complete_list)
print(time.time() - t, values[0:10], values[-10:-1])
t = time.time()
values = multi_process(complete_list)
print(time.time() - t, values[0:10], values[-10:-1])
# required for Windows:
if __name__ == '__main__':
main()
Prints:
4.181002378463745 [1236, 1491, 1464, 1477, 1494, 1472, 1410, 1450, 1502, 1537] [1485, 1513, 1513, 1501, 1283, 1538, 804, 1459, 1457]
1.325998067855835 [1236, 1491, 1464, 1477, 1494, 1472, 1410, 1450, 1502, 1537] [1485, 1513, 1513, 1501, 1283, 1538, 804, 1459, 1457]
Results
For CASE 1 the speedup was 3.86, for CASE 2 the speedup was only 3.14.