Yet Another Python Profiler


Keywords
python, thread, multithread, asyncio, gevent, profiler, asgi, asynchronous, coroutine, cpu, greenlet, multi-threaded-applications, multithreading, performance, profile, profilers
License
MIT
Install
pip install yappi==0.96

Documentation

yappi

Yappi

A tracing profiler that is multithreading, asyncio and gevent aware.

FreePalestine.Dev

From the river to the sea, Palestine will be free

Highlights

  • Fast: Yappi is fast. It is completely written in C and lots of love and care went into making it fast.
  • Unique: Yappi supports multithreaded, asyncio and gevent profiling. Tagging/filtering multiple profiler results has interesting use cases.
  • Intuitive: Profiler can be started/stopped and results can be obtained from any time and any thread.
  • Standards Compliant: Profiler results can be saved in callgrind or pstat formats.
  • Rich in Feature set: Profiler results can show either Wall Time or actual CPU Time and can be aggregated from different sessions. Various flags are defined for filtering and sorting profiler results.
  • Robust: Yappi has been around for years.

Motivation

CPython standard distribution comes with three deterministic profilers. cProfile, Profile and hotshot. cProfile is implemented as a C module based on lsprof, Profile is in pure Python and hotshot can be seen as a small subset of a cProfile. The major issue is that all of these profilers lack support for multi-threaded programs and CPU time.

If you want to profile a multi-threaded application, you must give an entry point to these profilers and then maybe merge the outputs. None of these profilers are designed to work on long-running multi-threaded applications. It is also not possible to profile an application that start/stop/retrieve traces on the fly with these profilers.

Now fast forwarding to 2019: With the latest improvements on asyncio library and asynchronous frameworks, most of the current profilers lacks the ability to show correct wall/cpu time or even call count information per-coroutine. Thus we need a different kind of approach to profile asynchronous code. Yappi, with v1.2 introduces the concept of coroutine profiling. With coroutine-profiling, you should be able to profile correct wall/cpu time and call count of your coroutine. (including the time spent in context switches, too). You can see details here.

Installation

Can be installed via PyPI

$ pip install yappi

OR from the source directly.

$ pip install git+https://github.com/sumerc/yappi#egg=yappi

Examples

A simple example:

import yappi

def a():
    for _ in range(10000000):  # do something CPU heavy
        pass

yappi.set_clock_type("cpu") # Use set_clock_type("wall") for wall time
yappi.start()
a()

yappi.get_func_stats().print_all()
yappi.get_thread_stats().print_all()
'''

Clock type: CPU
Ordered by: totaltime, desc

name                                  ncall  tsub      ttot      tavg      
doc.py:5 a                            1      0.117907  0.117907  0.117907

name           id     tid              ttot      scnt        
_MainThread    0      139867147315008  0.118297  1
'''

Profile a multithreaded application:

You can profile a multithreaded application via Yappi and can easily retrieve per-thread profile information by filtering on ctx_id with get_func_stats API.

import yappi
import time
import threading

_NTHREAD = 3


def _work(n):
    time.sleep(n * 0.1)


yappi.start()

threads = []
# generate _NTHREAD threads
for i in range(_NTHREAD):
    t = threading.Thread(target=_work, args=(i + 1, ))
    t.start()
    threads.append(t)
# wait all threads to finish
for t in threads:
    t.join()

yappi.stop()

# retrieve thread stats by their thread id (given by yappi)
threads = yappi.get_thread_stats()
for thread in threads:
    print(
        "Function stats for (%s) (%d)" % (thread.name, thread.id)
    )  # it is the Thread.__class__.__name__
    yappi.get_func_stats(ctx_id=thread.id).print_all()
'''
Function stats for (Thread) (3)

name                                  ncall  tsub      ttot      tavg
..hon3.7/threading.py:859 Thread.run  1      0.000017  0.000062  0.000062
doc3.py:8 _work                       1      0.000012  0.000045  0.000045

Function stats for (Thread) (2)

name                                  ncall  tsub      ttot      tavg
..hon3.7/threading.py:859 Thread.run  1      0.000017  0.000065  0.000065
doc3.py:8 _work                       1      0.000010  0.000048  0.000048


Function stats for (Thread) (1)

name                                  ncall  tsub      ttot      tavg
..hon3.7/threading.py:859 Thread.run  1      0.000010  0.000043  0.000043
doc3.py:8 _work                       1      0.000006  0.000033  0.000033
'''

Different ways to filter/sort stats:

You can use filter_callback on get_func_stats API to filter on functions, modules or whatever available in YFuncStat object.

import package_a
import yappi
import sys

def a():
    pass

def b():
    pass

yappi.start()
a()
b()
package_a.a()
yappi.stop()

# filter by module object
current_module = sys.modules[__name__]
stats = yappi.get_func_stats(
    filter_callback=lambda x: yappi.module_matches(x, [current_module])
)  # x is a yappi.YFuncStat object
stats.sort("name", "desc").print_all()
'''
Clock type: CPU
Ordered by: name, desc

name                                  ncall  tsub      ttot      tavg
doc2.py:10 b                          1      0.000001  0.000001  0.000001
doc2.py:6 a                           1      0.000001  0.000001  0.000001
'''

# filter by function object
stats = yappi.get_func_stats(
    filter_callback=lambda x: yappi.func_matches(x, [a, b])
).print_all()
'''
name                                  ncall  tsub      ttot      tavg
doc2.py:6 a                           1      0.000001  0.000001  0.000001
doc2.py:10 b                          1      0.000001  0.000001  0.000001
'''

# filter by module name
stats = yappi.get_func_stats(filter_callback=lambda x: 'package_a' in x.module
                             ).print_all()
'''
name                                  ncall  tsub      ttot      tavg
package_a/__init__.py:1 a             1      0.000001  0.000001  0.000001
'''

# filter by function name
stats = yappi.get_func_stats(filter_callback=lambda x: 'a' in x.name
                             ).print_all()
'''
name                                  ncall  tsub      ttot      tavg
doc2.py:6 a                           1      0.000001  0.000001  0.000001
package_a/__init__.py:1 a             1      0.000001  0.000001  0.000001
'''

Profile an asyncio application:

You can see that coroutine wall-time's are correctly profiled.

import asyncio
import yappi

async def foo():
    await asyncio.sleep(1.0)
    await baz()
    await asyncio.sleep(0.5)

async def bar():
    await asyncio.sleep(2.0)

async def baz():
    await asyncio.sleep(1.0)

yappi.set_clock_type("WALL")
with yappi.run():
    asyncio.run(foo())
    asyncio.run(bar())
yappi.get_func_stats().print_all()
'''
Clock type: WALL
Ordered by: totaltime, desc

name                                  ncall  tsub      ttot      tavg      
doc4.py:5 foo                         1      0.000030  2.503808  2.503808
doc4.py:11 bar                        1      0.000012  2.002492  2.002492
doc4.py:15 baz                        1      0.000013  1.001397  1.001397
'''

Profile a gevent application:

You can use yappi to profile greenlet applications now!

import yappi
from greenlet import greenlet
import time

class GreenletA(greenlet):
    def run(self):
        time.sleep(1)

yappi.set_context_backend("greenlet")
yappi.set_clock_type("wall")

yappi.start(builtins=True)
a = GreenletA()
a.switch()
yappi.stop()

yappi.get_func_stats().print_all()
'''
name                                  ncall  tsub      ttot      tavg
tests/test_random.py:6 GreenletA.run  1      0.000007  1.000494  1.000494
time.sleep                            1      1.000487  1.000487  1.000487
'''

Documentation

Related Talks

Special thanks to A.Jesse Jiryu Davis:

PyCharm Integration

Yappi is the default profiler in PyCharm. If you have Yappi installed, PyCharm will use it. See the official documentation for more details.