Memory profiling
The profiling techniques we have covered so far aim at finding the runtime bottlenecks. Let's briefly discuss memory profiling, another important aspect of profiling.
The memory_profiler package
For memory profiling, we will use a popular Python package called memory_profiler. It can be installed using pip. Here is how to install it on Linux from the command line:
$ pip install memory_profiler
The documentation highly recommends installing the psutils module. It also suggests that, in order for memory_profiler to work on Windows OS, you will need the psutil module. The psutil module can be installed using pip, as follows:
$ pip install psutil
Tip
For more information on memory_profiler, check out the following page: https://pypi.python.org/pypi/memory_profiler.
Just like line_profiler, the memory_profiler package uses the @profile decorator above the function name. Let's add the decorator @profile just above the generate_random_points function, and then run the memory profiler...
