Profiling your code line-by-line with line_profiler

Python's native cProfile module and the corresponding %prun magic break down the execution time of code function by function. Sometimes, we may need an even more fine-grained analysis of code performance with a line-by-line report. Such reports can be easier to read than reports from cProfile.

To profile code line-by-line, we need an external Python module named line_profiler. In this recipe, we will demonstrate how to use this module within IPython.

Getting ready

To install line_profiler, type conda install line_profiler in a Terminal.

How do to it...

We will profile the same simulation code as in the previous recipe, line-by-line.

1. First, let's import NumPy and the line_profiler IPython extension module that comes with the package:

   >>> import numpy as np
       %load_ext line_profiler

2. This IPython extension module provides an %lprun magic command to profile a Python function line-by-line. It works best when the function is defined...