Being a Programmer, one of the most important task is to analyse the code and optimize it , so that code should use less memory and time. One may think, finding a part of code which is taking up more space in memory is tedious task. But wait, fortunately in Python, we have different libraries for performance analysis, example: cProfile, line-profiler, guppy, memory-profiler. In this article, we’ll discuss memory-profiler library for monitoring memory consumption.

Module memory-profiler returns line-by-line analysis of memory consumption for python programs.

Installation

pip install -U memory_profiler

To get the line-by-line analysis follow below two steps:

1. import module.
2. use decorator @profile

In below example, we are calling my_func(), in which we are just did some calculation and stored in variable ‘a’ , and after that we deleted ‘a’. This program may not be related to real-world problem, but this code is enough to understand memory profiling.

from memory_profiler import profile
@profile
def my_func(x):
    a = [x]*(10**8)
    del a
my_func(20)

Output:

Line ##    Mem usage    Increment  Occurences   Line Contents
============================================================
     2     20.3 MiB     20.3 MiB           1   @profile
     3                                         def my_func(x):
     4    783.2 MiB    762.9 MiB           1       a = [x]*(10**8)
     5     20.3 MiB   -762.9 MiB           1       del a

#python #performance-analysis #optimization-techniques