The documentation for memory_profiler offers a simple example showing the changes in memory usage due to creating and deleting a list.
from memory_profiler import profile
@profile
def my_func():
a = [1] * (10 ** 6)
b = [2] * (2 * 10 ** 7)
del b
return a
if __name__ == '__main__':
my_func()
Which produces the following output:
Line # Mem usage Increment Line Contents
==============================================
3 @profile
4 5.97 MB 0.00 MB def my_func():
5 13.61 MB 7.64 MB a = [1] * (10 ** 6)
6 166.20 MB 152.59 MB b = [2] * (2 * 10 ** 7)
7 13.61 MB -152.59 MB del b
8 13.61 MB 0.00 MB return a
When b is deleted, its memory usage is freed up. I was playing around and changed the example to the following, where I also delete a.
from memory_profiler import profile
@profile
def my_func():
a = [1] * (10 ** 6)
b = [2] * (2 * 10 ** 7)
del a
del b
return None
if __name__ == '__main__':
my_func()
Here, however, deletion of a is not accompanied - as I would naively expect - by a decrease in memory usage. In fact, it seems as though nothing is happening.
Line # Mem usage Increment Line Contents
================================================
12 25.1 MiB 0.0 MiB @profile
13 def my_func():
14 32.7 MiB 7.6 MiB a = [1] * (10 ** 6)
15 185.3 MiB 152.6 MiB b = [2] * (2 * 10 ** 7)
16 185.3 MiB 0.0 MiB del a
17 32.7 MiB -152.6 MiB del b
18 32.7 MiB 0.0 MiB return None
What is going on?
del a does in fact not delete the object, it merely marks the object as not used but it will be deleted later, whenever the garbage collector deems it appropriate. b is big enough to make the garbage collector delete it immediately, but apparently not so for a