Repolode
Submit
pyutils

pyutils/line_profiler

Line-by-line profiler for Python that shows how much time each source line of a decorated function takes.

View on GitHub
Stars
3,252
Forks
143
Open issues
59
Last commit
20 days ago

What is it?

line_profiler measures execution time line by line inside Python functions. It is useful once function-level profiling has pointed to a slow function and you need to see which specific lines account for the time. Functions are marked with the @profile decorator, and when the script finishes it prints a per-line summary of hits, time, and percentage of total time.

The recommended workflow is to import profile from line_profiler, decorate the functions of interest, and set the LINE_PROFILE=1 environment variable before running the script. The package also ships kernprof, a command-line helper, which supports the older kernprof -lv workflow. Installation is through pip, and the project requires Python 3.10 or newer.

This repository is the maintained fork that the PyPI line_profiler package now points to, continuing work on the original project by rkern. Documentation covers basic usage, examples, and an FAQ.

Key features

  • 01Reports execution count, time, and percentage for each line of a decorated function.
  • 02Enables profiling with the @profile decorator and the LINE_PROFILE environment variable.
  • 03Includes the kernprof command-line helper.
  • 04Installable from PyPI with pip, with wheels for common platforms.
  • 05Optional IPython integration via the ipython extra.

Use cases

  • Finding which lines inside a slow Python function consume most of the runtime.
  • Profiling data processing or numerical scripts before optimizing them.
  • Checking the effect of a code change on specific lines of a hot function.

Quick start

  1. 1

    Install line_profiler with pip

    pip install line_profiler
  2. 2

    Import profile and decorate the function you want to profile

  3. 3

    Set LINE_PROFILE=1 and run your script as normal

Things to know

  • Requires Python 3.10 or newer; older Python versions need older releases.
  • Source installs may need a C compiler, and git checkouts also need Cython.

Topics

#python#profiling#performance#debugging#benchmarking#devtools

FAQ

What is line_profiler?
Line-by-line profiler for Python that shows how much time each source line of a decorated function takes. line_profiler measures execution time line by line inside Python functions. It is useful once function-level profiling has pointed to a slow function and you need to see which specific lines account for the time. Functions are marked with the @profile decorator, and when the script finishes it prints a per-line summary of hits, time, and percentage of total time.
Is line_profiler free to use? What license does it have?
line_profiler is open source under the Other license. Check the license terms before using it in commercial products.
Is line_profiler still maintained?
Actively maintained. The last commit to line_profiler was on Oct 18, 2026.
What can I use line_profiler for?
Finding which lines inside a slow Python function consume most of the runtime. Profiling data processing or numerical scripts before optimizing them. Checking the effect of a code change on specific lines of a hot function.
How do I install line_profiler?
Install line_profiler with pip: pip install line_profiler Import profile and decorate the function you want to profile Set LINE_PROFILE=1 and run your script as normal
Where can I get line_profiler?
The source code and installation instructions are on GitHub: https://github.com/pyutils/line_profiler

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects