Implementing a Python decorator to measure memory usage of a function

Python Decorator: Exercise-11 with Solution

Write a Python program that implements a decorator to measure the memory usage of a function.

Sample Solution:

Python Code:

import tracemalloc

def measure_memory_usage(func):
    def wrapper(*args, **kwargs):
        tracemalloc.start()

        # Call the original function
        result = func(*args, **kwargs)

        snapshot = tracemalloc.take_snapshot()
        top_stats = snapshot.statistics("lineno")

        # Print the top memory-consuming lines
        print(f"Memory usage of {func.__name__}:")
        for stat in top_stats[:5]:
            print(stat)

        # Return the result
        return result

    return wrapper

# Example usage
@measure_memory_usage
def calculate_factorial(n):
    if n == 0:
        return 1
    else:
        return n * calculate_factorial(n - 1)

# Call the decorated function
result = calculate_factorial(5)
print("Factorial:", result)

Sample Output:

Memory usage of calculate_factorial:
/tmp/sessions/aeafebf7cb54338f/main.py:29: size=2128 B, count=9, average=236 B
/tmp/sessions/aeafebf7cb54338f/main.py:8: size=1376 B, count=5, average=275 B
/tmp/sessions/aeafebf7cb54338f/main.py:10: size=416 B, count=1, average=416 B
Memory usage of calculate_factorial:
/tmp/sessions/aeafebf7cb54338f/main.py:29: size=2128 B, count=9, average=236 B
/tmp/sessions/aeafebf7cb54338f/main.py:8: size=1312 B, count=4, average=328 B
/usr/local/lib/python3.10/tracemalloc.py:226: size=880 B, count=3, average=293 B
/usr/local/lib/python3.10/tracemalloc.py:173: size=800 B, count=2, average=400 B
/usr/local/lib/python3.10/tracemalloc.py:533: size=568 B, count=1, average=568 B
Memory usage of calculate_factorial:
/tmp/sessions/aeafebf7cb54338f/main.py:29: size=2016 B, count=7, average=288 B
/usr/local/lib/python3.10/tracemalloc.py:67: size=1280 B, count=20, average=64 B
/usr/local/lib/python3.10/tracemalloc.py:535: size=1240 B, count=3, average=413 B
/usr/local/lib/python3.10/tracemalloc.py:193: size=1104 B, count=23, average=48 B
/usr/local/lib/python3.10/tracemalloc.py:226: size=880 B, count=3, average=293 B
Memory usage of calculate_factorial:
/tmp/sessions/aeafebf7cb54338f/main.py:29: size=1904 B, count=5, average=381 B
/usr/local/lib/python3.10/tracemalloc.py:558: size=1416 B, count=29, average=49 B
/usr/local/lib/python3.10/tracemalloc.py:67: size=1344 B, count=21, average=64 B
/usr/local/lib/python3.10/tracemalloc.py:535: size=1240 B, count=3, average=413 B
/usr/local/lib/python3.10/tracemalloc.py:226: size=832 B, count=2, average=416 B
Memory usage of calculate_factorial:
/tmp/sessions/aeafebf7cb54338f/main.py:29: size=1856 B, count=4, average=464 B
/usr/local/lib/python3.10/tracemalloc.py:67: size=1408 B, count=22, average=64 B
/usr/local/lib/python3.10/tracemalloc.py:535: size=1240 B, count=3, average=413 B
/usr/local/lib/python3.10/tracemalloc.py:558: size=1232 B, count=25, average=49 B
/usr/local/lib/python3.10/tracemalloc.py:226: size=880 B, count=3, average=293 B
Factorial: 24

Explanation:

In the above exercise -

The decorator function "measure_memory_usage()" measures the memory usage of a function using the tracemalloc module. Here's a brief explanation of the code:

• The measure_memory_usage decorator takes a function func as an argument.
• Within the wrapper function, tracemalloc.start() is called to start tracing memory allocations.
• The original function func is called with the provided arguments and the result is stored in the result variable.
• tracemalloc.take_snapshot() is used to take a snapshot of the memory allocation at that point in the code.
• The snapshot is used to retrieve memory statistics using snapshot.statistics("lineno"). This returns a list of statistics sorted by line number.
• The decorator then prints the top memory-consuming lines by iterating over the top_stats list and printing them.
• Finally, the original function result is returned.

In the example,

• The calculate_factorial function is decorated with @measure_memory_usage to measure its memory usage.
• The decorated function is called with an argument of 5 to calculate the factorial.
• The factorial result is printed along with the memory usage statistics obtained from the decorator.

Flowchart:

Have another way to solve this solution? Contribute your code (and comments) through Disqus.

Previous: Implementing a Python decorator for enforcing type checking on function arguments.
Next: Implementing a Python decorator for caching with expiration time in functions.