Visual profiler for Python

Nick Volynets

Last update: Jan 1, 2023

Related tags

Debugging Tools visualization javascript d3 python profiler stats developer-tools cpu-flame-graph

Overview

vprof

vprof is a Python package providing rich and interactive visualizations for various Python program characteristics such as running time and memory usage. It supports Python 3.4+ and distributed under BSD license.

The project is in active development and some of its features might not work as expected.

Screenshots

Contributing

All contributions are highly encouraged! You can add new features, report and fix existing bugs and write docs and tutorials. Feel free to open an issue or send a pull request!

Prerequisites

Dependencies to build vprof from source code:

Python 3.4+
pip
npm >= 3.3.12

npm is required to build vprof from sources only.

Dependencies

All Python and npm module dependencies are listed in package.json and requirements.txt.

Installation

vprof can be installed from PyPI

pip install vprof

To build vprof from sources, clone this repository and execute

python3 setup.py deps_install && python3 setup.py build_ui && python3 setup.py install

To install just vprof dependencies, run

python3 setup.py deps_install

Usage

vprof -c <config> <src>

<config> is a combination of supported modes:

c - CPU flame graph ⚠️ Not available for windows #62

Shows CPU flame graph for <src>.

p - profiler

Runs built-in Python profiler on <src> and displays results.

m - memory graph

Shows objects that are tracked by CPython GC and left in memory after code execution. Also shows process memory usage after execution of each line of <src>.

h - code heatmap

Displays all executed code of <src> with line run times and execution counts.

<src> can be Python source file (e.g. testscript.py) or path to package (e.g. myproject/test_package).

To run scripts with arguments use double quotes

vprof -c cmh "testscript.py --foo --bar"

Modes can be combined

vprof -c cm testscript.py

vprof can also profile functions. In order to do this, launch vprof in remote mode:

vprof -r

vprof will open new tab in default web browser and then wait for stats.

To profile a function run

from vprof import runner

def foo(arg1, arg2):
    ...

runner.run(foo, 'cmhp', args=(arg1, arg2), host='localhost', port=8000)

where cmhp is profiling mode, host and port are hostname and port of vprof server launched in remote mode. Obtained stats will be rendered in new tab of default web browser, opened by vprof -r command.

vprof can save profile stats to file and render visualizations from previously saved file.

vprof -c cmh src.py --output-file profile.json

writes profile to file and

vprof --input-file profile.json

renders visualizations from previously saved file.

Check vprof -h for full list of supported parameters.

To show UI help, press h when visualizations are displayed.

Also you can check examples directory for more profiling examples.

Testing

python3 setup.py test_python && python3 setup.py test_javascript && python3 setup.py e2e_test

License

BSD

Comments

`Exception happened during processing of request from ('127.0.0.1', 51937)`

Thanks for making this useful project available!

I pip installed it (Mac OS 10.11.3, Anaconda Python 3.5.1), and then tried running it. I get the messages saying that it's running the various profiling functionalities (depending on the flags), and then a blank browser tab opens pointing to localhost:8000. The message in the title then appears in the terminal.

Any ideas? Thanks again
bug

opened by arokem 23
Heatmap creator hangs if called with another profilers

I'm trying to profile http://scikit-image.org/docs/dev/auto_examples/filters/plot_inpaint.html . vprof {ch/hc/mh/hm} plot_inpaint.py works units of seconds, while vprof cmh ... doesn't finish even within 10mins.
bug

opened by soupault 17
Ability to choose profiled files

Description

I tested you package against django tests, after I saw what you were capable of doing using code heatmap, which renders great by the way. So I thought I would give it a try on my Django tests because selenium is taking long time, and a heatmap would definitely reveal what parts are slow so I can focus on them.

How to reproduce

$ vprof -c h "manage.py test --failfast"

Actual results

The code heatmap is only showing the manage.py file, would it be possible to add an option to monitor my test files (i.e. tests.py or any other file).

Expected results

I would like to have several files show up in the profiling result (that I pick beforehand)

Version and platform

vprof==0.36.1 MacOS 10.12.5 Python 3.5.2
feature request

opened by SebCorbin 14
Some imports in Python 3 break flame graph rendering
I ran vprof -c cmh -s "evaluate.py -o 7496a0471ba8259926b36028b79b5a0e62ecf03c -d 510 -f 260 -s 25" and no results were presented in any way. If I understand correctly there should appear new tabs in my browser but that didn't happen.

Also: I had to pass the hash without single quotation marks as vprof would put another pair around them. Is this intended?

Platform: python3.5:

Python 3.5.2 (default, Jun 28 2016, 08:46:01) [GCC 6.1.1 20160602] on linux

Chromium Version 53.0.2785.116 (64-bit)

I am running Antergos linux
bug
opened by sims1253 13

`python setup.py deps_install` doesn't complete properly

I'm not sure if it's deps_install that's failing, but whenever I open vprof in remote mode, http://localhost:/ shows a spinning circle that keeps on spinning forever. build_ui and install throw no noticeable errors. Here's the output from deps_install:

(mvenv)$ python setup.py deps_install
running deps_install
Requirement already satisfied (use --upgrade to upgrade): psutil>=3.4.2 in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from -r requirements.txt (line 1))
Requirement already satisfied (use --upgrade to upgrade): six>=1.10.0 in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from -r requirements.txt (line 2))
Requirement already satisfied (use --upgrade to upgrade): mock>=1.0.0 in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from -r dev_requirements.txt (line 1))
Requirement already satisfied (use --upgrade to upgrade): pylint>=1.5.4 in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from -r dev_requirements.txt (line 2))
Requirement already satisfied (use --upgrade to upgrade): six in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from pylint>=1.5.4->-r dev_requirements.txt (line 2))
Requirement already satisfied (use --upgrade to upgrade): astroid<1.5.0,>=1.4.5 in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from pylint>=1.5.4->-r dev_requirements.txt (line 2))
Requirement already satisfied (use --upgrade to upgrade): colorama in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from pylint>=1.5.4->-r dev_requirements.txt (line 2))
Requirement already satisfied (use --upgrade to upgrade): wrapt in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from astroid<1.5.0,>=1.4.5->pylint>=1.5.4->-r dev_requirements.txt (line 2))
Requirement already satisfied (use --upgrade to upgrade): lazy-object-proxy in /Users/ali/.pyenv/versions/mvenv/lib/python2.7/site-packages (from astroid<1.5.0,>=1.4.5->pylint>=1.5.4->-r dev_requirements.txt (line 2))
npm WARN package.json [email protected] No README data
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/karma/node_modules/socket.io/node_modules/engine.io/node_modules/engine.io-parser requires has-binary@'0.1.6' but will load
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/karma/node_modules/socket.io/node_modules/has-binary,
npm WARN unmet dependency which is version 0.1.7
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/karma/node_modules/socket.io/node_modules/socket.io-client/node_modules/engine.io-client requires component-emitter@'1.1.2' but will load
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/karma/node_modules/socket.io/node_modules/socket.io-client/node_modules/component-emitter,
npm WARN unmet dependency which is version 1.2.0
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/karma/node_modules/socket.io/node_modules/socket.io-adapter/node_modules/socket.io-parser requires debug@'0.7.4' but will load
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/karma/node_modules/socket.io/node_modules/debug,
npm WARN unmet dependency which is version 2.2.0
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/phantomjs-prebuilt/node_modules/request/node_modules/http-signature/node_modules/sshpk requires assert-plus@'^1.0.0' but will load
npm WARN unmet dependency /Users/ali/repo/vprof/node_modules/phantomjs-prebuilt/node_modules/request/node_modules/http-signature/node_modules/assert-plus,
npm WARN unmet dependency which is version 0.2.0

My self-test outputs are as follows:

(mvenv)$ python setup.py test
running test
testGetRunDispatcher (base_profile_test.BaseProfileUnittest) ... ok
testInit_RunObjFunction (base_profile_test.BaseProfileUnittest) ... ok
testInit_RunObjImportedPackage (base_profile_test.BaseProfileUnittest) ... ok
testInit_RunObjModule (base_profile_test.BaseProfileUnittest) ... ok
testInit_RunObjPackagePath (base_profile_test.BaseProfileUnittest) ... ok
testRun (base_profile_test.BaseProfileUnittest) ... ok
testRunAsFunction (base_profile_test.BaseProfileUnittest) ... ok
testRunAsModule (base_profile_test.BaseProfileUnittest) ... ok
testRunAsPackageInNamespace (base_profile_test.BaseProfileUnittest) ... ok
testRunAsPackagePath (base_profile_test.BaseProfileUnittest) ... ok
testGetPackageCode (base_profile_test.GetPackageCodeUnittest) ... ok
testAddCode (code_heatmap_test.CodeHeatmapCalculator) ... ok
testCalcHeatmap (code_heatmap_test.CodeHeatmapCalculator) ... ok
testInit (code_heatmap_test.CodeHeatmapCalculator) ... ok
testAddCode (memory_profile_test.CodeEventsTrackerUnittest) ... ok
testTraceMemoryUsage_EmptyEventsList (memory_profile_test.CodeEventsTrackerUnittest) ... ok
testTraceMemoryUsage_NormalUsage (memory_profile_test.CodeEventsTrackerUnittest) ... ok
testTraceMemoryUsage_OtherCode (memory_profile_test.CodeEventsTrackerUnittest) ... ok
testTraceMemoryUsage_SameLine (memory_profile_test.CodeEventsTrackerUnittest) ... ok
testTransformStats (runtime_profile_test.RuntimeProfileUnittest) ... ok

----------------------------------------------------------------------
Ran 20 tests in 0.022s

OK

> vprof-frontend@ test /Users/ali/repo/vprof
> karma start

19 05 2016 16:37:59.894:INFO [framework.browserify]: bundle built
19 05 2016 16:37:59.907:INFO [karma]: Karma v0.13.22 server started at http://localhost:9876/
19 05 2016 16:37:59.914:INFO [launcher]: Starting browser PhantomJS
19 05 2016 16:38:01.173:INFO [PhantomJS 2.1.1 (Mac OS X 0.0.0)]: Connected on socket /#-QYG9HhjEzuR9d5NAAAA with id 80466447
PhantomJS 2.1.1 (Mac OS X 0.0.0): Executed 3 of 3 SUCCESS (0.003 secs / 0.002 secs)
(mvenv)$ python setup.py e2e_test
running e2e_test
testRequest (code_heatmap_e2e.CodeHeatmapFunctionEndToEndTest) ... ok
testRequest (code_heatmap_e2e.CodeHeatmapImportedPackageEndToEndTest) ... ok
testRequest (code_heatmap_e2e.CodeHeatmapModuleEndToEndTest) ... ok
testRequest (code_heatmap_e2e.CodeHeatmapPackageAsPathEndToEndTest) ... ok
testRequest (memory_profile_e2e.MemoryProfileFunctionEndToEndTest) ... ok
testRequest (memory_profile_e2e.MemoryProfileImportedPackageEndToEndTest) ... ok
testRequest (memory_profile_e2e.MemoryProfileModuleEndToEndTest) ... ok
testRequest (memory_profile_e2e.MemoryProfilePackageAsPathEndToEndTest) ... ok
testRequest (runtime_profile_e2e.RuntimeProfileFunctionEndToEndTest) ... ok
testRequest (runtime_profile_e2e.RuntimeProfileImportedPackageEndToEndTest) ... ok
testRequest (runtime_profile_e2e.RuntimeProfileModuleEndToEndTest) ... ok
testRequest (runtime_profile_e2e.RuntimeProfilePackageAsPathEndToEndTest) ... ok

----------------------------------------------------------------------
Ran 12 tests in 6.114s

OK

I'm sure it's something really straightforward I've missed, because I didn't see any issues like this in other open (or closed) issues.

opened by alichaudry 9

Remove entries < 0.2% of total runtime

If you have a lot of files (say, thousands), many of them aren't going to account for more than 0.X% of total runtime. However they will be rendered in the result list, making the UI load for a very long time or crash.

This is a working PoC of removing any entry <0.2% of total runtime from the results prior to serving them to the UI.

Another approach may be to include only the top 99% (or whatever) time consumers, but this didn't work as well as the fixed threshold approach.

opened by jonashaag 6
Allow code heatmap to show time taken per line as well as execution count

Thanks for a great tool! This visual profiler is what I was looking for in Python.

I was surprised though that the code heatmap only shows the execution count for the colour highlighting, and for the mouseovers. For me it is far more interesting to know how long each line took, not just how many times it is run. Of course this is also somewhat in the flame graph, but not in the same format.

It would be great to be able to select what the code heatmap shows, effectively allowing the user to choose between the columns of line_profiler. This could be in the command-line options, but ideally could be in a drop-down menu in the code heatmap screen. I believe matlab's profiler offers something similar.

Thanks
feature request

opened by mdfirman 6

vprof crashes with a stacktrace on scripts with Python 3.6.1 32-bit and Windows

Description

vprof crashes with a stacktrace on scripts with Python 3.6.1 32-bit and Windows

How to reproduce

Use Windows 10.
Install Python 3.6.1 32 bit. (Also had issues with 64 bit.)
Type pip install vprof
create a script that says print("Hi") and save it as test.py
On the command line type vprof -c p test.py

Actual results

C:\>vprof -c p test.py
Running Profiler...
Traceback (most recent call last):
  File "c:\program files (x86)\python36-32\lib\runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "c:\program files (x86)\python36-32\lib\runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "C:\Program Files (x86)\Python36-32\Scripts\vprof.exe\__main__.py", line 9, in <module>
  File "c:\program files (x86)\python36-32\lib\site-packages\vprof\__main__.py", line 88, in main
    source, config, verbose=True)
  File "c:\program files (x86)\python36-32\lib\site-packages\vprof\runner.py", line 78, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "c:\program files (x86)\python36-32\lib\site-packages\vprof\base_profiler.py", line 162, in run
    return dispatcher()
  File "c:\program files (x86)\python36-32\lib\site-packages\vprof\base_profiler.py", line 71, in multiprocessing_wrapper
    process.start()
  File "c:\program files (x86)\python36-32\lib\multiprocessing\process.py", line 105, in start
    self._popen = self._Popen(self)
  File "c:\program files (x86)\python36-32\lib\multiprocessing\context.py", line 223, in _Popen
    return _default_context.get_context().Process._Popen(process_obj)
  File "c:\program files (x86)\python36-32\lib\multiprocessing\context.py", line 322, in _Popen
    return Popen(process_obj)
  File "c:\program files (x86)\python36-32\lib\multiprocessing\popen_spawn_win32.py", line 65, in __init__
    reduction.dump(process_obj, to_child)
  File "c:\program files (x86)\python36-32\lib\multiprocessing\reduction.py", line 60, in dump
    ForkingPickler(file, protocol).dump(obj)
AttributeError: Can't pickle local object 'run_in_another_process.<locals>.multiprocessing_wrapper.<locals>.remote_wrapper'

C:\>Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "c:\program files (x86)\python36-32\lib\multiprocessing\spawn.py", line 99, in spawn_main
    new_handle = reduction.steal_handle(parent_pid, pipe_handle)
  File "c:\program files (x86)\python36-32\lib\multiprocessing\reduction.py", line 82, in steal_handle
    _winapi.PROCESS_DUP_HANDLE, False, source_pid)
OSError: [WinError 87] The parameter is incorrect

Expected results

Not a stack trace.

Version and platform

Python 3.6.1 (v3.6.1:69c0db5, Mar 21 2017, 17:54:52) [MSC v.1900 32 bit (Intel)] on win32

bug

opened by pvcraven 5

SIGPROF does not work on Windows

Python 3.5.1 (v3.5.1:37a07cee5969, Dec  6 2015, 01:54:25) [MSC v.1900 64 bit (AMD64)] on win32

PS C:\Users\Admin\Downloads\trufont> vprof -c c Lib/trufont
Running FlameGraphProfiler...
Traceback (most recent call last):
  File "C:\Python35\Scripts\vprof-script.py", line 9, in <module>
    load_entry_point('vprof==0.34', 'console_scripts', 'vprof')()
  File "c:\python35\lib\site-packages\vprof\__main__.py", line 88, in main
    source, config, verbose=True)
  File "c:\python35\lib\site-packages\vprof\runner.py", line 78, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "c:\python35\lib\site-packages\vprof\flame_graph.py", line 169, in run
    prof = run_dispatcher()
  File "c:\python35\lib\site-packages\vprof\flame_graph.py", line 142, in run_as_package
    with _StatProfiler() as prof:
  File "c:\python35\lib\site-packages\vprof\flame_graph.py", line 29, in __enter__
    signal.signal(signal.SIGPROF, self.sample)
AttributeError: module 'signal' has no attribute 'SIGPROF'

bug known issue

opened by adrientetar 5

Don't suppress ImportErrors.

This is a tool for python programmers. We know what import errors are and they give us some idea what we're doing wrong. Replacing them with a vague error message doesn't help.

opened by ryneeverett 5
Pass back return value from the profiled function

vprof.runner.run(...) ignored return value of the function profiled. Would it be possible to pass it back?

(I would like to submit pull request but it actually is not as straightforward as I thought :/)
feature request

opened by eudoxos 3
Latest `highlight.js` node module breaks the build

Description

Fails to build with latest hightlight.js node module.

How to reproduce

Get the latest highlight.js and run python setup.py build_ui.

Fix

I looked into the commits over on highlight.js and found this commit stating

github-gist has been removed in favor of github [Jan Pilzer][]

If I update highlight.css and change

@import url('node_modules/highlight.js/styles/github-gist.css');

to

@import url('node_modules/highlight.js/styles/github.css');

everything works.

opened by JesseBuesking 0
How to use on Windows

On Windows, I did pip3 install vprof, but that didn't create any vprof.bat nor vprof.exe anywhere, so when I try something like vprof -c <config> <src> in the command line (as suggested by the readme), I get "'vprof' is not recognized as an internal or external command, operable program or batch file."

Should I add something specific to my environment variables?

opened by page200 1

Profiling command-line based scripts that change terminal settings results in a termios error

Description

Profiling a CLI script results in this error:

Traceback (most recent call last):
  File "/usr/local/bin/vprof", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.9/site-packages/vprof/__main__.py", line 83, in main
    program_stats = runner.run_profilers(source, config, verbose=True)
  File "/usr/local/lib/python3.9/site-packages/vprof/runner.py", line 73, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "/usr/local/lib/python3.9/site-packages/vprof/base_profiler.py", line 172, in run
    return self.profile()
  File "/usr/local/lib/python3.9/site-packages/vprof/flame_graph.py", line 167, in profile_module
    return base_profiler.run_in_separate_process(self._profile_module)
  File "/usr/local/lib/python3.9/site-packages/vprof/base_profiler.py", line 79, in run_in_separate_process
    raise exc
termios.error: (19, 'Operation not supported by device')

How to reproduce

run vprof -c ch Test.py inside this folder.

Actual results

This happens after a while

Traceback (most recent call last):
  File "/usr/local/bin/vprof", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.9/site-packages/vprof/__main__.py", line 83, in main
    program_stats = runner.run_profilers(source, config, verbose=True)
  File "/usr/local/lib/python3.9/site-packages/vprof/runner.py", line 73, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "/usr/local/lib/python3.9/site-packages/vprof/base_profiler.py", line 172, in run
    return self.profile()
  File "/usr/local/lib/python3.9/site-packages/vprof/flame_graph.py", line 167, in profile_module
    return base_profiler.run_in_separate_process(self._profile_module)
  File "/usr/local/lib/python3.9/site-packages/vprof/base_profiler.py", line 79, in run_in_separate_process
    raise exc
termios.error: (19, 'Operation not supported by device')

Expected results

(without profiler)
Screenshot 2021-11-24 at 10 39 15 AM (side note: ctrl+z to quit Test.py)

Version and platform

vprof 0.38
MacOs 10.14.6
Screenshot 2021-11-24 at 10 41 48 AM

opened by lomnom 0

object of type 'int' has no len()

Description

run command

vprof -c cmh test.py

Error message :

Running MemoryProfiler...
Building prefix dict from the default dictionary ...
Loading model from cache /tmp/jieba.cache
Loading model cost 0.488 seconds.
Prefix dict has been built successfully.
/home/binnz/miniconda3/envs/transformer-tf/lib/python3.8/site-packages/torch/distributed/distributed_c10d.py:144: UserWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
  warnings.warn("torch.distributed.reduce_op is deprecated, please use "
Running FlameGraphProfiler...
Traceback (most recent call last):
  File "/home/binnz/miniconda3/envs/transformer-tf/bin/vprof", line 8, in <module>
    sys.exit(main())
  File "/home/binnz/miniconda3/envs/transformer-tf/lib/python3.8/site-packages/vprof/__main__.py", line 83, in main
    program_stats = runner.run_profilers(source, config, verbose=True)
  File "/home/binnz/miniconda3/envs/transformer-tf/lib/python3.8/site-packages/vprof/runner.py", line 73, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "/home/binnz/miniconda3/envs/transformer-tf/lib/python3.8/site-packages/vprof/base_profiler.py", line 172, in run
    return self.profile()
  File "/home/binnz/miniconda3/envs/transformer-tf/lib/python3.8/site-packages/vprof/flame_graph.py", line 167, in profile_module
    return base_profiler.run_in_separate_process(self._profile_module)
  File "/home/binnz/miniconda3/envs/transformer-tf/lib/python3.8/site-packages/vprof/base_profiler.py", line 79, in run_in_separate_process
    raise exc
TypeError: object of type 'int' has no len()

Version and platform

Linux binnz-MS-7B89 5.4.0-66-generic #74~18.04.2-Ubuntu SMP Fri Feb 5 11:17:31 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux

vprof version: 0.38 python version: 3.8.8

opened by binnz 0

Fails with "IndexError: list index out of range" when running with -h (heatmap)

Description

I'm not really sure where it's coming from, but I'm getting the following error when I run the profiler in heatmap mode (with the -h flag).

How to reproduce

Haven't been able to find out exactly what causes it yet. I just ran my script as vprof -c h "/path/to/script.py --some_param" < input_file.txt

Actual results

Fails with

Traceback (most recent call last):
  File "/home/ldorigo/.local/bin/vprof", line 8, in <module>
    sys.exit(main())
  File "/home/ldorigo/.local/lib/python3.9/site-packages/vprof/__main__.py", line 83, in main
    program_stats = runner.run_profilers(source, config, verbose=True)
  File "/home/ldorigo/.local/lib/python3.9/site-packages/vprof/runner.py", line 73, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "/home/ldorigo/.local/lib/python3.9/site-packages/vprof/base_profiler.py", line 172, in run
    return self.profile()
  File "/home/ldorigo/.local/lib/python3.9/site-packages/vprof/code_heatmap.py", line 217, in profile_module
    return base_profiler.run_in_separate_process(self._profile_module)
  File "/home/ldorigo/.local/lib/python3.9/site-packages/vprof/base_profiler.py", line 79, in run_in_separate_process
    raise exc
IndexError: list index out of range

Expected results

Makes a line heatmap :)

Version and platform

Python 3.9.1 on linux, vprof 0.38

opened by ldorigo 1

[Help] How to test twisted code

Description

I have sample example, when i run, but an error occurred.

from twisted.internet import reactor, task
from vprof import runner


def ppp():
    print('called')


def profile():
    reactor.callLater(3, reactor.stop)
    t = task.LoopingCall(ppp)
    t.start(0.5)
    reactor.run()


if __name__ == '__main__':
    runner.run(profile, 'cmhp')

Traceback (most recent call last):
  File "/home/kevin/workspaces/develop/python/crawlerstack/crawlerstack_proxypool/demo.py", line 19, in <module>
    runner.run(profile, 'cmhp')
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/vprof/runner.py", line 91, in run
    run_stats = run_profilers((func, args, kwargs), options)
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/vprof/runner.py", line 73, in run_profilers
    run_stats[option] = curr_profiler.run()
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/vprof/base_profiler.py", line 172, in run
    return self.profile()
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/vprof/flame_graph.py", line 172, in profile_function
    result = self._run_object(*self._run_args, **self._run_kwargs)
  File "/home/kevin/workspaces/develop/python/crawlerstack/crawlerstack_proxypool/demo.py", line 15, in profile
    reactor.run()
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/twisted/internet/base.py", line 1282, in run
    self.startRunning(installSignalHandlers=installSignalHandlers)
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/twisted/internet/base.py", line 1262, in startRunning
    ReactorBase.startRunning(self)
  File "/home/kevin/.virtualenvs/crawlerstack_proxypool-VZQM9sgQ/lib/python3.7/site-packages/twisted/internet/base.py", line 765, in startRunning
    raise error.ReactorNotRestartable()
twisted.internet.error.ReactorNotRestartable

Version and platform

vprof 0.38
Linux 5.4.70-amd64-desktop #1 SMP Wed Oct 14 15:24:23 CST 2020 x86_64 GNU/Linux
Python 3.7.3
twisted 20.3.0

question

opened by whg517 0

Releases(v0.38)

v0.38(Feb 29, 2020)
Drop Python 2 support

Bug fixes and performance improvements

Source code(tar.gz)
Source code(zip)
v0.37.6(Jun 24, 2018)
Fixed

Fix Markdown rendering on PyPI.

Source code(tar.gz)
Source code(zip)
v0.37.5(Jun 24, 2018)
Fixed

pip install vprof ImportError: No module named download (#84).

Source code(tar.gz)
Source code(zip)
v0.37.4(Dec 10, 2017)
Fixed

Prevent crash on unexpected type rear (#73).

Performance regression during path calculation.

Source code(tar.gz)
Source code(zip)
v0.37.3(Oct 31, 2017)
Fixed

Flame graph zoom rendering.

Source code(tar.gz)
Source code(zip)
v0.37.2(Oct 12, 2017)
Fixed

Pass back return value from the profiled function (#69).

Source code(tar.gz)
Source code(zip)
v0.37.1(Sep 26, 2017)
Fixed

Code heatmap displays all code except code from standard library now (#66).

Source code(tar.gz)
Source code(zip)
v0.37(Jul 20, 2017)
vprof 0.37 has been released.

New features

Now you can see run time distribution for modules under "Inspected modules".

Also you can get info about current tab by pressing the "?" button in the top right corner.

Bug fixes and performance improvements.

Source code(tar.gz)
Source code(zip)
v0.36.1(May 1, 2017)
Fixed

vprof crashes with a stacktrace on scripts with Python 3.6.1 32-bit and Windows (#64)

Source code(tar.gz)
Source code(zip)
v0.36(Apr 9, 2017)

vprof 0.36 has been released.

New features

Code heatmap shows run times for individual lines

Now you can see run time distribution over executed lines on code heatmap tab. The feature is experimental and will be improved in future versions.

Source code(tar.gz)
Source code(zip)
v0.35(Jan 12, 2017)

vprof 0.35 is ready!

New features

Consistent flame graph coloring

Now every function in flame graph has it's own color allowing to explore function call patterns.
Source code(tar.gz)
Source code(zip)
v0.34(Nov 17, 2016)
vprof 0.34 has been released!

Thanks to all contributors!

New features

Render visualizations from file

vprof is able to save profile stats to file and render visualizations from previously saved file now.

vprof -c cmh src.py --output-file profile.json

writes profile to file and

vprof --input-file profile.json

renders visualizations from previously saved file.

Please note that stats rendering is supported for the same version only (e.g. vprof 0.35 won't render stats from 0.34).

Flame graph uses statistical profiler

Flame graph now uses statistical profiler instead of cProfile.

"Profiler" tab

New "Profiler" tab shows how long and how long often various parts of the programs are executed.

Other changes

-s flag has been removed from CLI.

Please, check `vprof -h' or README.md for more info.
Source code(tar.gz)
Source code(zip)
v0.33(Sep 21, 2016)
Hooray, vprof 0.33 has been released!

New features

"Objects left in memory" table

Shows objects that are tracked by CPython garbage collector and left in memory after code execution.

Improvements

Faster memory stats rendering.

More accurate memory usage measurement.

All data between client and server is gzipped now.

Minor UI adjustments.

Source code(tar.gz)
Source code(zip)
v0.32(Jun 30, 2016)
Various UI improvements.

Source code(tar.gz)
Source code(zip)
v0.31(Jun 12, 2016)
Various bug fixes and usability improvements.

Source code(tar.gz)
Source code(zip)
v0.3(May 19, 2016)
Add support for profiling single functions.

Add support for profiling Python packages.

Improve UI performance.

Bug fixes and other performance improvements.

Source code(tar.gz)
Source code(zip)
v0.22(Mar 2, 2016)
Bug fixes and performance improvements.

Source code(tar.gz)
Source code(zip)
v0.21(Feb 22, 2016)
Handle scripts with arguments properly.

Windows compatibility fixes.

Source code(tar.gz)
Source code(zip)
v0.2(Feb 10, 2016)
New code heatmap tab.

Memory graph redesign.

Bug fixes and performance improvements.

Source code(tar.gz)
Source code(zip)

Owner

Nick Volynets

GitHub

printstack is a Python package that adds stack trace links to the builtin print function, so that editors such as PyCharm can link you to the source of the print call.

printstack is a Python package that adds stack trace links to the builtin print function, so that editors such as PyCharm can link to the source of the print call.

101 Aug 26, 2022

Visual profiler for Python

Related tags

Overview

vprof

Screenshots

Contributing

Prerequisites

Dependencies

Installation

Usage

Testing

License

Comments

Description

How to reproduce

Actual results

Expected results

Version and platform

Description

How to reproduce

Actual results

Expected results

Version and platform

Description

How to reproduce

Fix

Description

How to reproduce

Actual results

Expected results

Version and platform

Description

Version and platform

Description

How to reproduce

Actual results

Expected results

Version and platform

Description

Version and platform

Releases(v0.38)

v0.38(Feb 29, 2020)

v0.37.6(Jun 24, 2018)

Fixed

v0.37.5(Jun 24, 2018)

Fixed

v0.37.4(Dec 10, 2017)

Fixed

v0.37.3(Oct 31, 2017)

Fixed

v0.37.2(Oct 12, 2017)

Fixed

v0.37.1(Sep 26, 2017)

Fixed

v0.37(Jul 20, 2017)

New features

v0.36.1(May 1, 2017)

Fixed

v0.36(Apr 9, 2017)

New features

Code heatmap shows run times for individual lines

v0.35(Jan 12, 2017)

New features

Consistent flame graph coloring

v0.34(Nov 17, 2016)

New features

Render visualizations from file

Flame graph uses statistical profiler

"Profiler" tab

Other changes

v0.33(Sep 21, 2016)

New features

"Objects left in memory" table

Improvements

v0.32(Jun 30, 2016)

v0.31(Jun 12, 2016)

v0.3(May 19, 2016)

v0.22(Mar 2, 2016)

v0.21(Feb 22, 2016)

v0.2(Feb 10, 2016)