Yet another introduction to the Python environment for beginners.

1. Python IntroductionEnrico Franchi (efranchi@ce.unipr.it)

2. We will perhaps eventually be writing only small
modules that are identified by name as they are used
to build larger ones, so that devices like indentation,
rather than delimiters, might become feasible for
expressing local structure in the source language.
Modules (identified by name and composed)
Indentation to express local structure
2
Donald E. Knuth,
Structured Programming with go to Statements, 1974

3. 1989
Started
Im
plem
entation1991
FirstPublicVersion
1994
StableRelease
2000
Python
2.0
2003
Python
2.3
2008
Python
3.0
Python Timeline
3

4. 1989
Started
Im
plem
entation1991
FirstPublicVersion
1994
StableRelease
2000
Python
2.0
2003
Python
2.3
2008
Python
3.0
Python Timeline
3
1991
Started
Im
plem
entation
1995
StableRelease
2000
Java
1.3
2004
Java
1.5
2006
Java
1.6
JAVA

5. Python 2 vs. 3
In order to greatly improve the language, GvR decided that a
newer major version that could break compatibility was needed
Python 3 was released in 2008 with the idea that the transition
would last at least 7 years (i.e., > 2015)
Now the “standard” version of Python is Python 2.7 (released in
2010). The world is mostly with Python 2.7
Most libraries are now Python 3 compatible and it is possible to
write code that works on both versions
In the future, Python 3 will be the standard
4

6. Why learning a new language
Because it makes easier to perform some tasks
To land a (better) job
To work with some other software/in some environment
To learn a new way of thinking
5

7. Language Categories
6
Declarative Imperative
Functional
Haskell, OCaml
Erlang, Lisp, Scheme
Logic
Prolog, Mercury, SQL
Imperative
Pascal, C
Object Oriented
Java, Ruby, C++,
Python, Smalltalk

8. Hybrid Languages
7
C++ Templates
C++
Object Oriented
Part
C++
Functions
Declarative Imperative
Functional Imperative
Object Oriented

9. “Functional” C++
8
template <int N>
struct Factorial
{
enum { value = N * Factorial<N - 1>::value };
};
template <>
struct Factorial<0>
{
enum { value = 1 };
};
int main() {
int num = Factorial<4>::value;
return 0;
}
fun fact 0 = 1
| fact n = n * fact (n - 1)
SML
f (n) =
1 n = 0
n⋅ f (n −1) otherwise
⎧
⎨
⎩

10. More “functional” C++
9
#include <iostream>
#include <algorithm>
#include <iterator>
#include <vector>
#include <functional>
int op_increase (int i) { return ++i; }
int main () {
const int foo[] = {10, 20, 30, 40, 50};
const size_t foo_size = sizeof(foo)/sizeof(int);
std::vector<int> bar(foo_size);
std::transform (foo, foo+foo_size, bar.begin(), op_increase);
// bar: 11 21 31 41 51
std::copy(bar.begin(), bar.end(),
std::ostream_iterator<int>(std::cout, ", "));
return 0;
}

11. Python
10
Python
Functional
Imperative
Object Oriented
“Functional” parts
of Python come
from the true
“everything is an
object” philosophy

12. Everything is an Object
In Python everything is an object
Values, Instances, Functions, ... even classes
11
def fact(n):
# I hate naive factorials
if n == 0:
return 1
else:
return n * fact(n-1)
alias_for_fact = fact
values = map(fact, [1, 2, 3, 4])
print values
# [1, 2, 6, 24]
class Person(object):
def __init__(self, name):
self.name = name
crew = map(Person, ["Malcom Reynolds", "Zoe Washburne",
"Kaylee Frye", "Jayne Cobb",
'Hoban "Wash" Washburne'])

13. Dynamic Typing Static Typing
Python, Ruby Haskell, SML, Java, (C++)
Perl, Javascript, PHP, Bash C
Strong
Typing
Weak
Typing
Strong
Typing
Weak
Typing
SM
L
H
askell
Java
C
++
C
Python
Typing
12

14. CPython
javac
Java Source Code
JVM
Python Source Code
Java Bytecode Python Bytecode
Execution
13

15. CPython
javac
Java Source Code
JVM
Python Source Code
Java Bytecode Python Bytecode
Execution
13

16. Interactive Shell
14
A read–eval–print loop (REPL) is
a simple, interactive computer
programming environment
Started with Lisp, most “good”
languages have it (Scala,
Haskell, Ruby, Perl, Scheme, ...)
It is useful to develop
interactively small (and not so
small) bits of code, and
especially to do scientific/
algorithmic computations

17. Why learning a new language
Because it makes easier to perform some tasks
To land a (better) job
To work with some other software/in some environment
To learn a new way of thinking
15

18. To learn a new way of thinking...
A language that does not affect the way you think about
programming, is not worth knowing.
Alan Perlis, 1966 Turing Award
16

19. READABILITY
GvR
Zen of Python
PEP8
Readability Counts
A program is read many
more times than it is
written.
Optimizing for readability
actually makes sense.
Python is built to be
highly readable (semantic
whitespace, low clutter,
no boilerplate)
17

20. 18
>>> import this
The Zen of Python, by Tim Peters
Beautiful is better than ugly.
Explicit is better than implicit.
Simple is better than complex.
Complex is better than complicated.
Flat is better than nested.
Sparse is better than dense.
Readability counts.
Special cases aren't special enough to break the rules.
Although practicality beats purity.
Errors should never pass silently.
Unless explicitly silenced.
In the face of ambiguity, refuse the temptation to guess.
There should be one-- and preferably only one --obvious way to do it.
Although that way may not be obvious at first unless you're Dutch.
Now is better than never.
Although never is often better than *right* now.
If the implementation is hard to explain, it's a bad idea.
If the implementation is easy to explain, it may be a good idea.
Namespaces are one honking great idea -- let's do more of those!

21. Why learning a new language
Because it makes easier to perform some tasks
To land a (better) job
To work with some other software/in some environment
To learn a new way of thinking
19

22. What helps...
Dynamic, High-Level: few lines of code make “lots” of stuff
Often “higher level” than design-patterns: several dp are from trivial
or even more verbose/less general
Almost no boilerplate: write what you mean
Highly readable: people strive to write elegant code
Small but very well thought set of primitive concepts
(Python “time-machine”)
Meta-programming: even higher level of abstraction, minor lines of
code and/or more general solutions (e.g., ORM)
20

23. ... and libraries
But mostly, a simple language encourages developers to
write libraries for various tasks
Lots of academics and companies wrote libraries for
almost every conceivable thing
PyPI (31113 libraries alone, plus all the project not indexed
but in GitHub, Bitbucket, Google Code or Sourceforge)
More on that later...
21

24. Why learning a new language
Because it makes easier to perform some tasks
To land a (better) job
To work with some other software/in some environment
To learn a new way of thinking
22

25. Python Community
Pycon US, EuroPython
Pycon.it, Python Italia A.P.S.
Widely used in the open source community: several key
projects are written in Python, and it is a default
component for several Linux distributions
Friendly and supportive community (very good social
capital), some “huge” names are Italians (and can help)
23

26. 24

27. Python Jobs
Learning new (good) languages is always good, even if you
are not going to use them (remember Perlis)
Python Paradox (see Graham, )
Several companies use Python
And often they are the “right” companies to work with
Job demand higher than job offer
25

28. Companies
26
ITA Software
Heroku, eBay, Microsoft,
Canonical, Spotify,
RedHat, Amazon, Netflix,
Adroll, Rackspace, Box,
Dropbox, Facebook,
Guidebook, Oracle,
Reddit, ....

29. Why learning a new language
Because it makes easier to perform some tasks
To land a (better) job
To work with some other software/in some
environment
To learn a new way of thinking
27

30. Scripting 3rd Party
28
SIMULIA Abaqus FEA
Blender, Cinema 4D,
Lightwave, Houdini, Maya,
modo, MotionBuilder,
Softimage, Nuke, GIMP,
Inkscape, Scribus and Paint
Shop Pro, GNU Debugger,
ArcGIS, QGIS
UNIX

31. Python Concurrency
Threads (bad solution, as always)
Twisted (asynchronous programming)
gevent (“light” threading for massive concurrency)
multiprocess (yeah, it’s UNIX!)
MPI, CUDA, etc...
29

32. Python and the Web
Full-stack frameworks (Django, Web2Py, Pylons, ...)
MVC, Component Based
Multi-DB ORM – URL Dispatcher – Templating System
Light-weight frameworks (Flask)
WSGI (simple web-server interface language)
30

33. Django
One of the most widely used
platform in the world
Pintertest, Instagram,
Washington Post,
ORM, URL Dispatcher, Web
Templating System
Automatic “Admin” generation
from the Model
Automatic translation from web
forms to entities
Built-in internationalization
Buil-in caching system
Extensible (built-in) authentication
system
Auto RSS & Atom generation
Built-in commenting system
Semi-auto Google Sitemap
Generation
Mitigation system for several
security issues (SQL Injection,
Cross-site scripting, ...)
31

34. Django Example
32
from django.db import models
class Poll(models.Model):
question = models.CharField(max_length=200)
pub_date = models.DateTimeField('date published')
class Choice(models.Model):
poll = models.ForeignKey(Poll)
choice_text = models.CharField(max_length=200)
votes = models.IntegerField(default=0)
>>> Poll.objects.all()
[<Poll: What's up?>]
# Django provides a rich database lookup API that's entirely driven by
# keyword arguments.
>>> Poll.objects.filter(id=1)
[<Poll: What's up?>]
>>> Poll.objects.filter(question__startswith='What')
[<Poll: What's up?>]

35. Django Admin
33

36. Flask Hello, World!
34
from flask import Flask
app = Flask(__name__)
@app.route("/")
def hello():
return "Hello World!"
if __name__ == "__main__":
app.run()
Extremely minimal web framework
Minimal memory requirements
Also comes with built-in server
Compact and easy to write
E.g., web interface to a Raspberry PI

37. HIS EX,
T E M P O
R NUMERIC PYTHON
• Numpy (low-level numerical computations) +
Scipy (lots of additional packages)
• Pandas (R-like)
• IPython (wonderfull command line interpreter) +
IPython Notebook (“Mathematica-like” interactive documents)
• HDF5 (PyTables, H5Py), Databases
• Specific libraries for machine learning, etc.
• General Purpose Object Oriented Programming

38. TOOLSCU
S E D

39. TOOLSCU
S E D

40. TOOLSCU
S E D

42. DENIQU
E
G U B E R
G R E N
Our Code
Numpy
Atlas/MKL
Improvements
Improvements
Algorithms are
fast because of
highly optimized
C/Fortran code
4 30 LOAD_GLOBAL 1 (dot)
33 LOAD_FAST 0 (a)
36 LOAD_FAST 1 (b)
39 CALL_FUNCTION 2
42 STORE_FAST 2 (c)
NUMPY STACK
c = a · b

43. ndarray
ndarray
Memory
behavior
shape, stride, flags
(i0, . . . , in 1) ! I
Shape: (d0, …, dn-1)
4x3
An n-dimensional array references some
(usually contiguous memory area)
An n-dimensional array has
property such as its shape or the
data-type of the elements containes
Is an object, so there is some
behavior, e.g., the def. of
__add__ and similar stuff
N-dimensional arrays are homogeneous

44. (i0, . . . , in 1) ! I
C-contiguousF-contiguous
Shape: (d0, …, dn)
IC =
n 1X
k=0
ik
n 1Y
j=k+1
dj
IF =
n 1X
k=0
ik
k 1Y
j=0
dj
Shape: (d0, …, dk ,…, dn-1)
Shape: (d0, …, dk ,…, dn-1)
IC = i0 · d0 + i14x3
IF = i0 + i1 · d1
ElementLayout
inMemory

45. Stride
C-contiguous F-contiguous
sF (k) =
k 1Y
j=0
dj
IF =
nX
k=0
ik · sF (k)
sC(k) =
n 1Y
j=k+1
dj
IC =
n 1X
k=0
ik · sC(k)
Stride
C-contiguousF-contiguous
C-contiguous
(s0 = d0, s1 = 1) (s0 = 1, s1 = d1)
IC =
n 1X
k=0
ik
n 1Y
j=k+1
dj IF =
n 1X
k=0
ik
k 1Y
j=0
dj

46. ndarray
Memory
behavior
shape, stride, flags
ndarray
behavior
shape, stride, flags
View View
View View
Views

49. C-contiguous
ndarray
behavior
(1,4)
Memory

50. C-contiguous
ndarray
behavior
(1,4)
Memory

51. ndarray
Memory
behavior
shape, stride, flags
matrix
Memory
behavior
shape, stride, flags
ndarray
matrix

52. BasicIndexing

53. AdvancedIndexing
Broadcasting!

54. AdvancedIndexing
Broadcasting!

55. AdvancedIndexing
Broadcasting!

56. AdvancedIndexing
Broadcasting!

57. AdvancedIndexing2

58. AdvancedIndexing2

59. AdvancedIndexing2

60. AdvancedIndexing2

61. AdvancedIndexing2

62. AdvancedIndexing2

63. Vectorize!
Don’t use explicit for loops unless you have to!