Introduction to Python for Scientific Computing

I introduce the Python programming language: main features, history, who uses it, and why it is a good choice for doing scientific computing.

Transcript

1. Introduction to Python for Scientific Computing Yoav Ram April 2019 1 Encyclopædia Britannica Online
2. What is Python? Python is a • Widely used • High-level • General-purpose • Interpreted • Dynamic Programming language Wikipedia
3. Design emphasizes code readability • Uncluttered visual layout (whitespaces…) • English keywords used where other languages use punctuation (and, or, not…) • Aims for simplicity and generality • The Python mantra: There should be one—and preferably only one— obvious way to do it • As opposed to the Perl and Ruby mantra: There's more than one way to do it Wikipedia
4. Example C void foo(int x) { if (-1 < x && x < 1) { bar(); baz(); } else { qux(x); foo(x - 1); } } Python def foo(x): if -1 < x < 1: bar() baz() else: qux(x) foo(x - 1)
5. Multiple programming paradigms • Object-oriented programming • Imperative programming • Functional programming • Procedural programming • Event driven programming • Asynchronous programming Wikipedia
6. Language features • Interpreted language • Dynamic type system (duck-typing) • Automatic memory management (GC) • Large and comprehensive standard library Wikipedia
7. Multi-platform • Interpreters available for many operating systems • Code can be executed on a wide variety of systems • Code can be packaged into stand-alone executable programs Wikipedia
8. Python culture • Free and open-source software • Community-based development model • Managed by the non-profit Python Software Foundation (PSF) • CPython is the reference implementation of Python • Other implementations: – IronPython for .NET framework, written in C# – Jython for Java framework – PyPy interpreter and JIT compiler, written in Python – MicroPython for microcontrollers and embedded systems – … Wikipedia
9. History of Python • Developed in 1989-91 by Guido van Rossum in the Netherlands • Python 2.0 released Oct 2000 • Many major new features: – cycle-detecting garbage collector – support for Unicode – shift to transparent and community-backed development • Python 3.0 released Dec 2008 – major backwards-incompatible release – many of major features backported to Python 2.6 and 2.7 • Python 3.7 released June 2018 • Python 3.8 alpha released; final version October 2019
10. Version history Python 1.0 - January 1994 Python 1.5 - December 31, 1997 Python 1.6 - September 5, 2000 Python 2.0 - October 16, 2000 Python 2.1 - April 17, 2001 Python 2.2 - December 21, 2001 Python 2.3 - July 29, 2003 Python 2.4 - November 30, 2004 Python 2.5 - September 19, 2006 Python 2.6 - October 1, 2008 Python 2.7 - July 3, 2010 Python 3.0 - December 3, 2008 Python 3.1 - June 27, 2009 Python 3.2 - February 20, 2011 Python 3.3 - September 29, 2012 Python 3.4 - March 16, 2014 Python 3.5 - September 13, 2015 Python 3.6 - December 16, 2016 Python 3.7 – June 27, 2018 Python 3.8 – October, 2019
11. Guido van Rossum • Python's principal author • Still has a central role in deciding the direction of Python development • Titled by the Python community: Benevolent Dictator for Life (BDFL) • Employed by Google 2005-2012 • Spent half his time developing Python • Since 2013 works for Dropbox • Spends half his time developing Python… Wikipedia Homepage
12. Why Python? I used Matlab. Now I use Python. by Steve Tjoa Why use Python for scientific computing?
13. Python is Free
14. Gratis: Free as in Beer • MATLAB is expensive – Individuals: $2,350 – Academia: $550 – Personal: $95 – Student: $29-55 – Batteries (toolboxes…) not included • Python is totally free – Batteries included (NumPy, SciPy…) MathWorks Pricing
15. Libre: Free as in Speech • MATLAB source code is closed and proprietary – You cannot see the code – You cannot change the code – You can participate in the discussion as a client • Python source code is open – You can see, you can change, you can contribute code and documentation (python, numpy) – You can participate in the discussion as a peer (python, numpy)
16. Python is a general-purpose language
17. Python is used for: • Scientific computing • Enterprise software • Web design • Back-end • Front-end • Everything in between
18. Python is used at Google, Rackspace, Microsoft, Intel, Walt Disney, MailChimp, twilio, Bank of America, Facebook, Instagram, HP, Linkedin, Elastic, Mozilla, YouTube, ILM, Thawte, CERN, Yahoo!, NASA, Trac, Civilization IV, reddit, LucasFilms, D-Link, Phillips, AstraZeneca, Applied Materials, KLA-Tencor, Nova, Lam Research, Marvell https://us.pycon.org/2016/sponsors/ https://www.python.org/about/quotes/ https://en.wikipedia.org/wiki/Python_%28programming_language%29#Use https://en.wikipedia.org/wiki/List_of_Python_software
19. How Dropbox Did It and How Python Helped Rian Hunter, a Dropbox Engineer presented at PyCon 2011: • 99.9 % of code in Python. • Server backend, desktop client, website controller logic, API backend, and analytics. • Run on a single code base using Python: Windows, Mac, Linux using tools like PyObjs, WxPython, types, py2exe, py2app, PyWin32. • Python helped iterate fast through error cases they experienced on the wide variety of platforms they support. • Use C for inner loops - optimizing CPU is easy. • Custom memory allocator - optimizing memory is harder. See more at highscalability
20. Success story: Philips • Semiconductor manufacturing facility in Fishkill, NY • In 1997 they started redesigning the system architecture • Python was suggested • Concern if a scripting language is suitable for the bulk of the code • Some favored significant portions of code in C++ • Everybody seemed to have a preference that wasn't Python • After much discussion, Python prevailed • The project was a huge success • Rebuilt 8 years of software development effort in less <2 years with a smaller team • Success attributed largely to Python - it is very easy to develop code quickly: 1. Python requires less supporting code – less boilerplate 2. Python speeds the development cycle – no compilation 3. Python facilitates debugging – even without using debugger • Later on, moving the system from OS/2 to Linux required almost no effort Michael Muller, https://www.python.org/about/success/philips/
21. Python is portable More or less same code runs on Windows, Linux, macOS, and any platform with a Python interpreter Python for "other" platforms
22. Python syntax is beautiful
23. Python syntax is beautiful Once you get over the use of meaningful whitespace, you realize how much it makes sense. Famous entrepreneur and investor Paul Graham: You spend more time reading code than writing it. You push blobs of source code around the way a sculptor does blobs of clay. So a language that makes source code ugly is maddening to an exacting programmer, as clay full of lumps would be to a sculptor. The Python Paradox, by Paul Graham
24. Python is inherently object- oriented
25. Almost everything is an object • strings, lists, dictionaries, tuples, functions, classes, and more • The implied usefulness is that these things each have their own members and methods that encapsulate its functionality and information • Strong polymorphism
26. Python is high level, easy to learn, and fast to develop So is MATLAB
27. Python has many cool features Stack overflow: Hidden features of Python
28. XKCD 353
29. Python is fast enough Written in C (and some Fortran) Easy to wrap more C Easy to parallelize JIT compiler over LLVM Benchmark Game | NumFocus Benchmarks
30. Python is popular and has a great community
31. Great community • Programmers • Scientists • Mathematicians • Engineers
32. Easy to find help on the Internet • Python is 6th most popular tag on StackOverflow • MATLAB is 60th most popular tag on StackOverflow stackoverflow.com/tags, May 2018
33. Active community • 3rd most active repositories on GitHub after Java (incl. Android) and JavaScript (incl. node.js) • ~27-fold more than MATLAB • As of May 2018 • See breakdown at githut
34. Python has a lot of great libraries
35. Many new libraries released every month During 2015, > 1,500 new packages released every month to PyPI. See more stats at PyGarden/stats.
36. Python can do nearly everything MATLAB can do With libraries like NumPy, SciPy, Matplotlib, IPython/Jupyter, Scikit-image, Scikit-learn, and more
37. Differential equations x = scipy.integrate.odeint(f, w0, t, …) plot(t, x[:, 0]) plot(t, x[:, 1])
38. Curve fitting params, cov = scipy.optimize.curve_fit( f=logistic, xdata=t, ydata=y, p0=(1, 10, 1)) N0=1.512, K=8.462, r=0.758
39. Optimization res = scipy.optimize.minimize_scalar( f, method="bounded", bounds=[8, 16]) fun: -0.23330441717143405 message: 'Solution found.' nfev: 9 status: 0 success: True x: 11.706005
40. Image analysis segmented = image > threshold dilated = scipy.ndimage.generic_filter(segmented, max) labels = skimage.measure.label(dilated) image labels
41. Machine learning knn = sklearn.neighbors.KNeighborsClassifier() knn.fit(X_train, y_train) knn.predict(X_test) Accuracy: 0.9
42. Demand & supply of Python programmers is high
43. Coding Dojo
44. Phillip Gou @ CACM
45. But why Python? Guido was reading some Monty Python's Flying Circus sketches and thought Python would be a cool name Python 2 FAQ
46. Thanks for listening 46 yoav@yoavram.com @yoavram github.com/yoavram python.yoavram.com Death to the Stock Ph