sipro 1.0.1

A very simple profiler for Python.

Nothing to see here...

...just a very simple python profiler

Examples

With context manager

Code

# Use default profiler
from sipro import default_profiler as profiler

# Create your own profiler
from sipro import Profiler

my_profiler = Profiler()

with profiler('my loop'):
    for i in range(1000):
        pass

with profiler('nested parent'):
    for i in range(50):
        pass

    with profiler('nested child 1'):
        for i in range(50):
            pass

    with profiler('nested child 2'):
        for i in range(50):
            pass

print(profiler)

Output

100.0% (100.0%)
my loop:       37.4% (37.4%)
nested parent: 17.6% (17.6%)
  nested child 1: 14.5% (2.5%)
  nested child 2: 11.3% (2.0%)
  ...:            74.2% (13.0%)
...:           45.0% (45.0%)

With functions

Code

from sipro import default_profiler as profiler


@profiler.wrap
def inherit_name():
    for i in range(1000):
        pass

from functools import partial

@partial(profiler.wrap, name='custom_name')
def dont_use_fn_name():
    for i in range(1000):
        pass

    # Functions also nest properly
    inherit_name()

dont_use_fn_name()

print(profiler)

Output

100.0% (100.0%)
custom_name: 71.4% (71.4%)
  inherit_name: 42.2% (30.1%)
  ...:          57.8% (41.3%)
...:         28.6% (28.6%)

Output

There are two ways of getting the output:

1. Cast to string str(profiler) - Output same as above
2. Get dict, by profiler.dict() - Structure:

   {
       'sum': <time in seconds, float>,
       'children': {
           '<child_name>': {
               ...child values
           }
       }
   }