This article will introduce you to a method of measuring the execution time of your Python code snippets. We will be using an in-built Python library timeit. This module provides a simple way to find the execution time of small bits of Python code.
What is the use of Timeit?
Well, how about using a simple time module? Just save the time before and after the execution of code and subtract them! But this method is not precise as there might be a background process momentarily running which disrupts the code execution and you will get significant variations in the running time of small code snippets. Timeit runs your snippet of code millions of times (default value is 1000000) so that you get the statistically most relevant measurement of code execution time! Timeit is pretty simple to use and has a command-line interface as well as a callable one.
Python Timeit() Syntax:
Syntax: timeit.timeit(stmt, setup, timer, number)
Parameter:
- stmt which is the statement you want to measure; it defaults to ‘pass’.
- setup which is the code that you run before running the stmt; it defaults to ‘pass’.
We generally use this to import the required modules for our code.- timer which is a timeit.Timer object; it usually has a sensible default value so you don’t have to worry about it.
- the number which is the number of executions you’d like to run the stmt.
Returns the number of seconds it took to execute the code.
A simple example of timeit() function
Python3
# testing timeit()import timeit# code snippet to be executed only oncemysetup = "5+5"print(timeit.timeit(mysetup)) |
Output:
0.0237445189995924
Timing Multiple Lines in Python Code
There are two ways to run numerous lines of code in timeit.timeit(), a semicolon, or by saving the code wrapped in triple quotes as a string.
Example 1: Using a semicolon
Python3
# testing timeit()import timeit# code snippet to be executedprint("The time taken is ",timeit.timeit (stmt='x=15;y=15;sum=x+y')) |
Output:
The time taken is 0.06830515900219325
Example 2: Using triple Quotes
The program is pretty straightforward. All we need to do is to pass the code as a string to the timeit.timeit() function. It is advisable to keep the import statements and other static pieces of code in the setup argument.
Python3
# importing the required moduleimport timeit# code snippet to be executed only oncemysetup = "from math import sqrt"# code snippet whose execution time is to be measuredmycode = '''def example(): mylist = [] for x in range(100): mylist.append(sqrt(x))'''# timeit statementprint(timeit.timeit(setup=mysetup, stmt=mycode, number=10000)) |
Output:
Note: Pay attention to the fact that the output is the execution time of the number times iterations of the code snippet, not the single iteration. For a single iteration exec. time, divide the output time by a number.
0.0024192919954657555
Example 3: Using timeit.repeat()
Let’s see another practical example in which we will compare two searching techniques, namely, Binary search and Linear search. Also, here we demonstrate two more features, timeit.repeat function and call the functions already defined in our program.
Python3
# importing the required modulesimport timeit# binary search functiondef binary_search(mylist, find): while len(mylist) > 0: mid = (len(mylist))//2 if mylist[mid] == find: return True else if mylist[mid] < find: mylist = mylist[:mid] else: mylist = mylist[mid + 1:] return False# linear search functiondef linear_search(mylist, find): for x in mylist: if x == find: return True return False# compute binary search timedef binary_time(): SETUP_CODE = '''from __main__ import binary_searchfrom random import randint''' TEST_CODE = '''mylist = [x for x in range(10000)]find = randint(0, len(mylist))binary_search(mylist, find)''' # timeit.repeat statement times = timeit.repeat(setup=SETUP_CODE, stmt=TEST_CODE, repeat=3, number=10000) # printing minimum exec. time print('Binary search time: {}'.format(min(times)))# compute linear search timedef linear_time(): SETUP_CODE = '''from __main__ import linear_searchfrom random import randint''' TEST_CODE = '''mylist = [x for x in range(10000)]find = randint(0, len(mylist))linear_search(mylist, find) ''' # timeit.repeat statement times = timeit.repeat(setup=SETUP_CODE, stmt=TEST_CODE, repeat=3, number=10000) # printing minimum exec. time print('Linear search time: {}'.format(min(times)))if __name__ == "__main__": linear_time() binary_time() |
Output:
The output of the above program will be the minimum value in the list times.
- strate below how you can utilize the command line
- timeit.repeat() function accepts one extra argument, repeat. The output will be a list of the execution times of all code runs repeated a specified no. of times.
- In the setup argument, we passed:
from __main__ import binary_search from random import randint
- This will import the definition of the function binary_search, already defined in the program, and the random library function randint.
- As expected, we notice that the execution time of binary search is significantly lower than linear search!
Example 4: default_timer()
In this example, we find the difference in time between the start and the end.
Python3
import timeitimport randomdef fun(): return random.randint(100, 800)start = timeit.default_timer()print("The start time is :", start)fun()print("The difference of time is :", timeit.default_timer() - start) |
Example 5
Finally, we demonstrate below how you can utilize the command line interface of Timeit module:
Here we explain each term individually:
So, this was a brief yet concise introduction to Timeit module and its practical applications. It’s a pretty handy tool for Python programmers when they need a quick glance at the execution time of their code snippets.
