One, two, buckle my shoe

Programming is more than just typing, and takes more than talent. Producing software takes time, dedication, and teamwork. This talk gives guiding points for those learning to program and a demonstration of setting up a Flask web app.

Transcript

1. ONE, TWO, BUCKLE MY SHOE Kat Chuang @katychuang PyGotham 2015
2. What this talk is about ▪ Compare experience of coding in different languages. – Python vs Haskell – Objected oriented vs Functional programming ▪ Share insights to things I’ve learned – Right-brained, visual approaches – It’s important to publish readable code – It’s possible to code with only text editor – no IDEs!
3. A little bit about me ▪ Pythonista – Research & data science – Web development ▪ NYC Python & NYC PyLadies organizer ▪ Now dabbling in Haskell too! – Functional Programming FTW! Want to learn data science? Take my online course! J https://teamtreehouse.com/signup_code/DrKat
4. Main Points 1. Plumbing is usually boring. 2. There's no I in team. 3. A year from now you'll wish you started today. 4. Demo
5. 1. PLUMBING IS USUALLY BORING. It’s not glamorous, but it’s necessary.
6. Python is an Object-Oriented language Object Oriented Programming (OOP) ▪ This is noun-orientation, objects are classes that have attributes. – Good when you have a fixed set of operations on things Real World Analogy ▪ Sneaker – Material – Size – Brand – calculatePrice – calculateMileage ▪ Sandal – Material – Size – Brand – calculatePrice – calculateMileage
7. Think more abstractly with functions ▪ Here are the pitfalls of OOP: – Adding new operations require editing many classes – Error prone ▪ Advice: write objected oriented code in the style of functions, then compose them As long as your items fit in the bins….
8. Example def my_greeting(): return ‘Hi’ def reverse_list(lst): lst = lst[::-1] def do_my_stuff(lst): print(my_greeting) reverse_list(lst)
9. Focusing on functions ▪ Write with functions ▪ Functions are easier to compose – Functions should not have side effects – same output every time – Functions should be limited in scope and functionality – keep it simple ▪ Example: def reversed(numbers): new_list = list(reversed(numbers)) return new_list
10. Recap: • Look at the pipe connections • It is like a built in unit test for type errors
11. 2. THERE'S NO I IN TEAM. Again, it’s not about the glamour.
12. Unreadable code slows you down… ▪ Handing off code to teammate or client and no documents ▪ Leaving a project for someone else taking over ▪ Coming back to code you wrote years ago ▪ Do you really want to spend hours figuring out if a bug is just a typo?
13. Which is easier to read? def reverse_list(lst): lst = lst[::-1] def swap_ends(lst): lst[0], lst[-1] = lst[-1], lst[0] def greeting(): return ‘Hi’ def do_my_stuff(lst): reverse_list(lst) swap_ends(lst) lst.insert(0, greeting)) def mutate(lst): lst = lst[::-1] lst[0], lst[-1] = lst[-1], lst[0] lst.insert(0, ‘Hi’) if isinstance(lst[0], str): lst[0] = lst[0][::-1] elif isinstance(lst[0], (float, int)): lst[0] = lst[0] // 2 return lst
14. How to write more readable code ▪ PEP 8 style guide https://www.python.org/dev/peps/pep-0008/ – Guidelines for indentation, line lengths, commenting ▪ Keep your code simple, consider it poetry ▪ Writing simple functions helps you manage complexity better ▪ Functions should not have side effects – it should give the same output every time ▪ Perform assertions of types
15. Publish readable code You’ll thank yourself later
16. 3. A YEAR FROM NOW YOU'LL WISH YOU STARTED TODAY. Baby steps all the way
17. Remember… ▪ Working on any large project is similar to gardening. ▪ Aim for done first, not perfect. ▪ You don’t need fancy tools
18. What can you accomplish in a year?
19. All you need is a text editor Lots of low cost options ▪ Sublime Text http://www.sublimetext.com/ ▪ Vim https://github.com/macvim-dev/macvim ▪ Emacs http://emacsformacosx.com/ ▪ PyCharm https://www.jetbrains.com/pycharm/ Combine a text editor with code searching functionality ▪ Silver Searcher http://geoff.greer.fm/ag/
20. Where to learn programming? ▪ Books – Learn Python the Hard Way http://learnpythonthehardway.org/book/ ▪ Videos – Treehouse http://teamtreehouse.com ▪ Local User Groups – NYC Python Study Groups and Office Hours http://learn.nycpython.org
21. Recap • Plumbing is usually boring. But when it works, we don’t worry. • Coding is not the same as regurgitating document specifications. It’s about building something usable. • It’s important to have the ability to reason about types.
22. 4. EXAMPLE APP Flask Web Development
23. Let’s make a dashboard of the latest Yeezy sneakers sold on eBay… Why sneaker analytics? http://fivethirtyeight.com/features/you-see-sneakers-these-guys-see-hundreds-of-millions-in-resale-profit/
24. Steps 1. Connect the eBay API 2. Get the Image URLs 3. Get the prices and calculate average 4. Send these values to the template(s) 5. Ta-Da!
25. 1. Connect the eBay API ▪ pip install ebaysdk ▪ Connect the API api = Connection(appid=settings.api_secret, config_file=None) response = api.execute('findItemsAdvanced', {'keywords’: ‘Yeezy’}) items = response.reply.searchResult.item
26. 2. Get the Image Urls def get_thumb(item): try: assert(type(item.galleryPlusPictureURL)) return item.galleryPlusPictureURL except: try: assert(type(item.galleryURL)) return item.galleryURL except: return ””
27. 3. Get the prices and calculate average def get_price(item): return item.sellingStatus.currentPrice.value def avg(my_list): stats = pd.DataFrame(my_list, columns=['Name', 'Img', 'Price’, 'Link']) stats[['Price']] = stats[['Price']].astype(float) avg_shoe = stats[['Price']].mean() total = stats[['Price']].sum()
28. 4. Send these values to the template(s) shoe = [title=item.title, currentPrice=get_price(item), galleryURL=get_thumb(item))] context = {“shoe”: shoe} return render_template("index.html", **context)
29. QUESTIONS? Contact me @katychuang http://twitter.com/katychuang http://github.com/katychuang hello@katychuang.com