Faster Python apps
Nuitka
Nuitka is a Python compiler written in Python, so it should work in all the different versions of Python (i.e., 2.6 through 3.9). Because Nuitka compiles Python code to C code, it has two requirements: A C compiler and Python must be installed on the machine where the application is being compiled.
Nuitka works on Linux, macOS, and Windows. For Linux, the GCC compiler is the default (5.1 or later). The Clang compiler is used on macOS, and the MinGW64 or Visual Studio 2019 compilers can be used for Windows. The Nuitka User Manual [5] discusses situations in which you might need to use a higher version of Python.
To install Nuitka and compile a test project (e.g., mytest.py
), simply enter:
python -m pip install nuitka python -m nuitka mytest.py
The Nuitka compiled program will be mytest.bin
in Linux, mytest
on macOS, and mytest.exe
in Windows.
I found that Nuitka had no problem with GUI libraries like tkinter, PySimpleGUI, and tk_tools.
Reusing Older Hardware
Python performance improvements with PyPy and Nuitka mean you can reuse some of the older Raspberry Pi and OpenWRT hardware (Figure 2). Unlike Nuitka, PyPy doesn't support the RPi.GPIO library, but a simple workaround is to shell out to the gpio
utility. An example to setup/write to/read from GPIO pin 7 with PyPy is shown in Listing 2.
Listing 2
PyPy and GPIO
§§noumber $ pypy3 Python 3.5.3 (7.0.0+dfsg-3, Mar 03 2019, 06:11:22) >>>> import os >>>> ret = os.system("gpio mode 7 output") >>>> ret = os.system("gpio write 7 1") >>>> ret = os.system("gpio read 7") 1
I had good success on a sailboat project that used a Raspberry Pi Zero W. With PyPy, I noticed some improved performance with the Python Bottle web framework. The control of the rudder was shelled out to the gpio utility.
Conclusion
A number of other options can make Python code run faster (e.g., Pyston and Cython), but I found PyPy and Nuitka to be the best supported.
Infos
- PyPy: https://www.pypy.org
- Nuitka: http://nuitka.net/
- PyPy package compatibility: http://packages.pypy.org
- PyPy3 plus virtualenv: https://techoverflow.net/2019/11/21/how-to-install-pypy3-on-the-raspberry-pi/
- Nuitka User Manual: https://nuitka.net/doc/user-manual.html#requirements
« Previous 1 2
Buy this article as PDF
(incl. VAT)