line_profiler

line_profiler is a pip package that can be used to profile python code, and get a result with line-by-line execution details.

Usage:

1. Create your python file with an annotated function that you’d like to profile

   import numpy as np
    
   @profile
   def BoxMuller(N):
       theta = np.random.uniform(low=0, high=2 * np.pi, size=N)
       E = np.random.exponential(scale=1, size=N)
       X = np.cos(theta) * np.sqrt(2 * E) # f(theta, E)
       return X
    
   if __name__ == '__main__':
       BoxMuller(10_000)

2. Run the file with kernprof:

   kernprof -l box_muller.py

   This will generate an lprof file that can be analyzed with line_profiler:

3. Obtain the metrics:

   python -m line_profiler box_muller.py.lprof
    
   Timer unit: 1e-06 s
    
   Total time: 0.077948 s
   File: box_muller.py
   Function: BoxMuller at line 4
    
   Line #      Hits         Time  Per Hit   % Time  Line Contents
   ==============================================================
        4                                           @profile
        5                                           def BoxMuller(N):
        6         1      11921.0  11921.0     15.3      theta = np.random.uniform(low=0, high=2 * np.pi, size=N)
        7         1      23561.0  23561.0     30.2      E = np.random.exponential(scale=1, size=N)
        8         1      42464.0  42464.0     54.5      X = np.cos(theta) * np.sqrt(2 * E) # f(theta, E)
        9         1          2.0      2.0      0.0      return X

Here’s an example without wrapping:

---

It can even be run in Jupyter notebooks with the use of the %line_profiler extension (source):

!pip install line_profiler
%load_ext line_profiler
 
def test_function():
    # some intense calculations
 
%lprun -f test_function test_function()