mbench 0.0.5

Simple Benchmarking Library.

mbench

---

Table of Contents

• mbench
  • Table of Contents
  • Installation
  • Usage
  • Caller Mode
  • Callee Mode
  • Using as a context manager
  • Using as a decorator
  • License

Installation

pip install mbench

Usage

from mbench import profileme
profileme(mode="caller")

def some_function():
    print("Hello")

some_function()

Hello
Function: some_function
  Duration: 0.000706 seconds
  CPU time: 0.000668 seconds
  Memory usage: 2.80 MB
  GPU usage: 0.00 MB
  I/O usage: 0.00 MB
  Avg Duration: 0.000527 seconds
  Avg CPU time: 0.000521 seconds
  Avg Memory usage: 0.35 MB
  Avg GPU usage: 0.00 MB
  Avg I/O usage: 0.00 MB
  Total calls: 8
-----------------------------
Profiling data saved to profiling_data.csv

Caller Mode

Functions you want to profile must

1. Be defined in the same module that the profileme function is being called.
2. Be called after profileme() is called.

Callee Mode

Functions you want to profile must

1. Be called in the same module that the profileme function is being called.
2. Be called after profileme() is called.

Using as a context manager

from mbench import profile_block

  def some_function():
        print("Hello")

with profileme():
    some_function()

Using as a decorator

from mbench import profile

@profile
def some_function():
    print("Hello")

License

mbench is distributed under the terms of the MIT License.