Profiling a program enables to know if the code is fast and which parts are slowing the execution time. By having these informations, it's easier to optimize the slow code. In python cProfile can be used to profile a code with:
import cProfile
cProfile.run(<Function to test or code to test>)CProfile is included in python by default, no installation are needed. It can also be used to profile a entire file by using this command:
python -m cProfile <PathFile>\<NameFile>.py
Profiling the code will sometimes gives too much information in the terminal and it will be impossible to see the complete output. There is a module to make it more user-friendly: snakeviz. To install it:
pip install snakeviz
Snakeviz is taking a .profile and convert it into a html file to make a GUI for the user to see the result of the profiling. To get a .profile file:
python -m cProfile -o <PathFile>\<NameOfNewFile>.profile <PathFile>\<NameFile>.py
Then, when the .profile has been generated:
python -m snakeviz <PathFile>\<NameOfNewFile>.profile
Doing this will open the browser and lead to a html page which will show a graphic and a array of the profiling. It's possible to sort the array.
The array can be sorted with these arguments : ncalls tottime percall cumtime percall. To notice : it is possible to click on a row to show its graphic and show what is slowing this part of code.