LineProfiler

This plugin exposes a simple interface to line_profiler and kernprof inside Sublime Text.

Installation

LineProfiler is available via Package Control, which is the easiest way to install it. Alternatively, you can download this repository and place it in your Packages directory.

Configuration

1. Install line_profiler and kernprof.py.
   • Easiest way to install is pip install --pre line_profiler.
   • Also installable by downloading from the line_profiler website.
2. Optional: Update LineProfiler's settings
   • Accessible at (Preferences > Package Settings > LineProfiler > “Settings - User”)

Available preferences are:

• kernprof: (Optional) path to the kerprof.py file you downloaded alongside the line_profiler module.
• pythonpath: (Optional) colon-separated list of paths that will be appended to the start of the the default PYTHONPATH.
• poll_timeout_seconds: (Optional) number of seconds to wait before killing the profiler. If you're profiling code that runs for longer than a minute, increase this value.
• poll_sleep_seconds: (Optional) number of seconds to sleep before polling the profiler for results. If you find the delay in getting profiler results too long, decrease this value.

Usage

Add the @profile annotation to any functions you'd like to profile. Then, run the plugin via the available bindings: * ctrl-alt-shift-p * right-click > “Run line_profiler” * Tools > Packages > “Run line_profiler”

If successful, a scratch buffer will appear with your results. If not, check the console (ctrl-`) to see any error messages.

The profiler runs in a background process, so Sublime will continue operating normally until a result is returned.

Example

Here's a simple way to try out LineProfiler and get familiar with the plugin.

Step 1: Open Sublime and write a simple Python file:

def foo():
  for i in xrange(1000):
    bar = i**2
    baz = i*i

if __name__ == '__main__':
  foo()

Sure, it doesn't do much, but it'll run. Be sure to tell Sublime that it's Python code, either by saving it as a .py file, or selecting Python in the bottom bar.

Step 2: Add the @profile decorator to the foo function, like so:

@profile
def foo():

(Everything else is unchanged.)

Step 3: Run LineProfiler, either from the Command Palette, the right-click menu, or the keyboard shortcut. The command is only available in Python mode, so check that first.

Step 4: Wait about a second, and then watch as the profile results appear in a new tab. If all went well, you should see the lines of the foo function annotated with their run counts, total time, and time per run:

function foo in file tmpev_rzr.py took 0.001248 s

Line #      Hits         Time  Per Hit   % Time  Line Contents
==============================================================
     2                                           @profile
     3                                           def foo():
     4      1001          372      0.4     29.8    for i in xrange(1000):
     5      1000          458      0.5     36.7      bar = i**2
     6      1000          418      0.4     33.5      baz = i*i

Some lines may be highlighted. These are “hot spots” in your code, which take up a disproportionate amount of time.

If you don't see anything after a second, check the console (ctrl-`) to see what went wrong. Often, the issue is a simple matter of setting the right paths in the user settings file.