Python courses

Introduction to programming using python and CS1 beginner training course


Télécharger Introduction to programming using python and CS1 beginner training course

★★★★★★★★★★3.5 étoiles sur 5 basé sur 1 votes.
Votez ce document:

Télécharger aussi :


Simple, not Simplistic

Squeezing the most from CS1 Python!

John M. Zelle, Ph.D.

Wartburg College

Outline

Motivation

Introduction to Python

Approaches to CS1

Python Resources

Conclusions  Questions?

Background

  Teaching since 1986

  CS1 languages: Pascal, C++, Java (also CS0 BASIC)

  Favorite class but   increasingly frustrating

  Students stopped "getting it"

  Student confusion, apathy, dropout

  Inability to complete simple programs

  Declining student evaluations

  Is it me?

Rethinking CS1

  Learning Challenges

  More material (software development, OOP, GUIs)

  Complex Languages (systems languages Ada, C++, Java)  Complex Environments

  Too much "magic"

  Teaching Challenges

  Recruiting Majors

  Serving Nonmajors

  Einstein: Make everything as simple as possible, but not  simpler.

The March of Progress (Cay Horstmann)

  C | Pascal  printf("%10.2f", x); | write(x:10:2) 

  C++

  cout << setw(10) << setprecision(2)

  << showpoint << x;

  Java

  .NumberFormat formatter  = .NumberFormat.getNumberInstance();  formatter.setMinimumFractionDigits(2);  formatter.setMaximumFractionDigits(2);  String s = formatter.format(x);  for (int i = s.length(); i < 10; i++)  .print(’ ’);

  .print(s);

Enter Python

  Python: A free, portable, dynamically-typed,  object-oriented scripting language

  Combines software engineering features of traditional systems languages with power and flexibility of scripting  languages

  Real world language

  Batteries included

  Note: Named after Monty Python’s Flying Circus

Why Use Python?

  Traditional languages (C++, Java) evolved for large-scale programming

  Emphasis on structure and discipline

  Simple problems != simple programs

  Scripting languages (Perl, Python, TCL) designed for simplicity and flexibility.

  Simple problems = simple, elegant solutions

  More amenable to experimentation and incremental development

  Python: Near ideal first language, useful throughout  curriculum

  We’ve used it in CS1 since 1998

First Program (Java Version)

  Assignment: Print "Hello CCSC" on screen

  public class Hello{

  public static void main(String [] args){  .println("Hello CCSC");

  }

  }

  Note: Must be in ""

First Program (Python Version)

  Assignment: Print "Hello CCSC" on screen

  print "Hello CCSC" 

 Or

  def main():  print "Hello CCSC"  main()

"Real" Program:

 #File:   # A simple program illustrating chaotic behavior.

 def main():

  print "This program illustrates a chaotic function"  x = input("Enter a number between 0 and 1: ")  for i in range(10):

  x = 3.9 * x * (1 - x)

  print x  main()

Example in IDLE

Basic Statements

  Output  print , , ,

 Note: all Python types have printable representations

  Simple Assignment

  =   myVar = oldValue * foo + skip 

  Simultaneous Assignment

  , , = , ,

  a,b = b,a 

  Assigning Input

  input()  myVar = input("Enter a number: ")  x,y = input("Enter the coordinates (x,y): ")

Example Program: Fibonacci

  #

  # This program computes the nth Fibonacci number

  n = input("Enter value of n ") 

  cur,prev = 1,1  for i in range(n-2):  cur,prev = prev+cur,cur  print "The nth Fibonacci number is", cur

Teaching Tip: Dynamic Typing

  Pluses

  less code  less upfront explanation

  eliminates accidental redeclaration errors

  Minuses

  typo on LHS of = creates new variable  allows variables to change type

  Bottom-line: I prefer dynamic types

  Many (most?) type errors are declaration errors

  Actual type errors are still detected

  Finding type errors goes hand-in-hand with testing

  Less student frustration

Teaching Tip: Indentation as Syntax

  Pluses

  less code clutter (; and {})  eliminates most common syntax errors  promotes and teaches proper code layout

  Minuses

  occasional subtle error from inconsistent spacing  will want an indentation-aware editor

  Bottom-line: Good Python editors abound.

  This is my favorite feature.

Numeric Types

  int: Standard 32 bit integer

  32 -3432 0

  long int: Indefinitely long integers  32L 9999999999999999

  floating-point: Standard double-precision float  3.14  2.57e-10  5E210  -3.64e+210

  complex: Double precision real and imaginary components

 2+3j  4.7J  -3.5 + 4.3e-4j

  User-defined types (operator overloading)

Numeric Operations

  Builtins

  +, -, *, /, %, **, abs(), round()

  Math Library

  pi, e, sin(), cos(), tan(), log(),  log10(), ceil(),

Example Numeric Program:

  #

  # Program to calculate real roots

  #  of a quadratic equation

  import math  a, b, c = input("Enter the coefficients (a, b, c): ") 

  discRoot = (b * b - 4 * a * c)  root1 = (-b + discRoot) / (2 * a)  root2 = (-b - discRoot) / (2 * a)  print "\nThe solutions are:", root1, root2

String Datatype

  String is an immutable sequence of characters

  Literal delimited by ’ or " or """

  s1 = ’This is a string’  s2 = "This is another"  s3 = "that’s one alright"  s4 = """This is a long string that  goes across multiple lines.

  It will have embedded end of lines"""

  Strings are indexed

  From the left starting at 0 or

  From the right using negative indexes

  A character is just a string of length 1

 String Operations

  >>>"Hello, " + " world!"

  ’Hello, world!’

  >>> "Hello" * 3

  ’HelloHelloHello’

  >>> greet = "Hello John"

  >>> print greet[0], greet[2], greet[4]

  H l o

  >>> greet[4:9]

  ’o Joh’

  >>> greet[:5]



  ’Hello’

  >>> greet[6:]

  ’John’

  >>> len(greet)

  10

Example Program: Month Abbreviation

  months = "JanFebMarAprMayJunJulAugSepOctNovDec" 

  n = input("Enter a month number (1-12): ")  pos = (n-1)*3  monthAbbrev = months[pos:pos+3]  print "The month abbreviation is", monthAbbrev+"."

More String Operations

  Interactive input  s = raw_input("Enter your name: ") 

  Looping through a string

  for ch in name:

  print ch 

  Type conversion

  to string

  >>> str(10)

  ’10’

 from string

  >>> eval(’10’)

  10  >>> eval(’3 + 4 * 7’)

    31

  Standard String Library (string)

  capitalize(s)  -- upper case first letter  capwords(s)  -- upper case each word  upper(s)  -- upper case every letter  lower(s)  -- lower case every letter 

  ljust(s, width)  -- left justify in width  center(s, width)  -- center in width  rjust(s, width)  -- right justify in width 

  count(substring, s)  -- count occurrences  find(s, substring)  -- find first occurrence  rfind(s, substring)  -- find from right end  replace(s, old, new) -- replace first occurrence 

  strip(s)  -- remove whitespace on both ends  rstrip(s) -- remove whitespace from end  lstrip(s) -- remove whitespace from front 

  split(s, char)  -- split into list of substrings  join(stringList) -- concatenate list into string

Example Programs: Text/ASCII Conversions

  # Converting from text to ASCII codes  message = raw_input("Enter message to encode: ") 

  print "ASCII Codes:"  for ch in message:  print ord(ch), 

  # Converting from ASCII codes to text

  import string  inString = raw_input("Enter ASCII codes: ") 

  message = ""  for numStr in string.split(inString):  message += chr(eval(numStr))  print "Decoded message:", message

String Formatting

  % operator inserts values into a template string (ala C printf)

  % ()

  "Slots" specify width, precision, and type of value

  %.

  Examples

  >>> "Hello %s %s, you owe %d" % ("Mr.", "X", 10000)

  ’Hello Mr. X, you owe 10000’

  >>> "ans = %8.3f" % 3.14159265

  ’ans =  3.142’  print "%10.2f" % x  # apparently, a throwback :-)

File Processing

  Opening a file

  syntax:  = open(, )  example: infile = open("", "r") 

  Reading from file

  syntax:  .read()

    .readline()  .readlines()  example: data = () 

  Writing to file

  syntax:  .write()  example: outfile.write(data)

Example Program: Username Creation

  Usernames are first initial and 7 chars of lastname (e.g.  jzelle).

  inf = open("", "r")  outf = open("", "w") 

  for line in inf:  first, last = line.split()  uname = (first[0]+last[:7]).lower()  outf.write(uname+’\n’) 

  inf.close()

  outf.close() 

  Note use of string methods (Python 2.0 and newer)

Functions

  Example:

  def distance(x1, y1, x2, y2):  # Returns dist from pt (x1,y1) to pt (x2, y2)  dx = x2 - x1  dy = y2 - y1  return (dx*dx + dy*dy) 

  Notes:

  Parameters are passed by value

  Can return multiple values

  Function with no return statement returns None

  Allows Default values

  Allows Keyword arguments

  Allows variable number of arguments

Teaching Tip: Uniform Memory Model

  Python has a single data model

  All values are objects (even primitive numbers)

  Heap allocation with garbage collection

  Assignment always stores a reference

  None is a special object (analogous to null)

  Pluses

  All assignments are exactly the same

  Parameter passing is just assignment

  Minuses

  Need to be aware of aliasing when objects are mutable

Decisions

  if temp > 90:

  print "It’s hot!" 

  if x <= 0:  print "negative"  else:

  print "nonnegative" 

  if x > 8:

  print "Excellent"  elif x >= 6:  print "Good"  elif x >= 4:  print "Fair"  elif x >= 2:  print "OK"  else:

  print "Poor"

Booleans in Python

  Traditional Python: Conditions return 0 or 1 (for false, true)

  As of Python 2.3 bool type: True, False

  All Python built-in types can be used in Boolean exprs

  numbers: 0 is False anything else is true  string: empty string is False, any other is true  None: False

  Boolean operators: and, or, not (short circuit, operational)

Loops

  For loop iterates over a sequence

  for in :

 

  sequences can be strings, lists, tuples, files, also user-defined classes  range function produces a numeric list  xrange function produces a lazy sequence

  Indefinite loops use while

  while :

 

  Both loops support break and continue

Lists: Dynamic Arrays

  Python lists are similar to vectors in Java

  dynamically sized  indexed (0..n-1) sequences

  But better..

  Heterogeneous

  Built into language (literals [])

  Rich set of builtin operations and methods

 Sequence Operations on Lists

  >>> x = [1, "Spam", 4, "U"]

  >>> len(x)

    4

  >>> x[3]

    ’U’

  >>> x[1:3]

  [’Spam’, 4]

  >>> x + x

  [1, ’Spam’, 4, ’U’, 1, ’Spam’, 4, ’U’]

  >>> x * 2

  [1, ’Spam’, 4, ’U’, 1, ’Spam’, 4, ’U’]

  >>> for i in x: print i,

  1 Spam 4 U

List are Mutable

  >>> x = [1, 2, 3, 4]

  >>> x[1] = 5

  >>> x

  [1, 5, 3, 4]



  >>> x[1:3] = [6,7,8]

  >>> x

  [1, 6, 7, 8, 4]

  >>> del x[2:4]

  >>> x

  [1, 6, 4]

List Methods

 myList.append(x)  -- Add x to end of myList  ()  -- Sort myList in ascending order  myList.reverse()  -- Reverse myList  myList.index(s)  -- Returns position of first x  myList.insert(i,x) -- Insert x at position i  myList.count(x)  -- Returns count of x  myList.remove(x)  -- Deletes first occurrence of x  (i)  -- Deletes and return ith element  x in myList  -- Membership check (sequences)

Example Program: Averaging a List

  def getNums():    nums = []    while True:    xStr = raw_input("Enter a number: ")    if xStr == "": break    nums.append(eval(xStr))    return nums

  def average(lst):

  sum = 0.0    for num in lst:    sum += num    return sum / len(lst) 

  data = getNums()  print "Average =", average(data)

Tuples: Immutable Sequences

  Python provides an immutable sequence called tuple

  Similar to list but:

  literals listed in () Aside: singleton (3,)  only sequence operations apply (+, *, len, in, iteration)  more efficient in some cases

  Tuples (and lists) are transparently "unpacked"

  >>> p1 = (3,4)

  >>> x1, y1 = p1

  >>> x1

  3

  >>> y1

  4

Dictionaries: General Mapping

  Dictionaries are a built-in type for key-value pairs (aka  hashtable)

  Syntax similar to list indexing

  Rich set of builtin operations

  Very efficient implementation

Basic Dictionary Operations

 >>> dict = { ’Python’: ’Van Rossum’, ’C++’:’Stroustrup’,  ’Java’:’Gosling’}

 >>> dict[’Python’]

’Van Rossum’

  >>> dict[’Pascal’] = ’Wirth’

 >>> ()

 [’Python’, ’Pascal’, ’Java’, ’C++’]

 >>> dict.values()

 [’Van Rossum’, ’Wirth’, ’Gosling’, ’Stroustrup’]

 >>> dict.items()

 [(’Python’, ’Van Rossum’), (’Pascal’, ’Wirth’), (’Java’,  ’Gosling’), (’C++’, ’Stroustrup’)]

More Dictionary Operations

 del dict[k]  -- removes entry for k  dict.clear()  -- removes all entries  dict.update(dict2)  -- merges dict2 into dict  dict.has_key(k)  -- membership check for k  k in dict    -- Ditto

 (k,d)  -- dict[k] returns d on failure  dict.setDefault(k,d) -- Ditto, also sets dict[k] to d 

Example Program: Most Frequent Words

  import string, sys 

  text = open([1],’r’).read()  text = text.lower()  for ch in string.punctuation:    text = text.replace(ch, ’ ’) 

  counts = {}  for w in text.split():    counts[w] = (w,0) + 1 

  items = [(c,w) for (w,c) in counts.items()]  ()

   items.reverse()

  for c,w in items[:10]:

  print w, c 

Python Modules

  A module can be:

  any valid source (.py) file  a compiled C or C++ file

  A single module can contain any number of structures

  Example: (GraphWin, Point, Line, Circle, color_rgb, )

  Locating modules

  Default search path includes Python lib and current directory

  Can be modified when Python starts or by program ()

  No naming or location restrictions

  Also supports directory structured packages

  from import *  from import *

Teaching Tip: Information Hiding

  In Python, Information hiding is by convention

  All objects declared in a module can be accessed by importers

  Names beginning with _ are not copied over in a from import *

  Pluses

  Makes independent testing of modules easier

  Eliminates visibility constraints (public, protected, private, static, etc.)

  Minuses

  Language does not enforce the discipline

  Bottom-line: Teaching the conventions is easier

  The concept is introduced when students are ready for it

 Simply saying "don’t do that" is sufficient (when grades are involved).

Python Classes: Quick Overview

  Objects in Python are class based (ala SmallTalk, C++,

  Java)

  Class definition similar to Java

  class :

   

  Class defines a namespace, but not a classic variable scope

  Instance variables qualified by an object reference

  Class variables qualified by a class or object reference

  Multiple Inheritance Allowed

Example: a generic multi-sided die

  from random import randrange 

    class MSDie:

  instances = 0  # Example class variable 

  def __init__(self, sides):

    self.sides = sides  self.value = 1

    MSDie.instances += 1

  def roll(self):  self.value = randrange(1, self.sides+1) 

  def getValue(self):   return self.value

Using a Class

  >>> from msdie import *

  >>> d1 = MSDie(6)

  >>> ()

  >>> d1.getValue()

  6

  >>> ()

  >>> d1.getValue()

  5

  >>> d1.instances

  1

  >>> MSDie.instances

  1

  >>> d2 = MSDie(13)

  >>> ()

  >>> d2.value

  7

  >>> MSDie.instances

  2

Example with Inheritance

    class SettableDie(MSDie):

  def setValue(self, value):    self.value = value

  ----------------------------------------------

  >>> import sdie

  >>> s = sdie.SettableDie(6)

  >>> s.value

  1

  >>> s.setValue(4)

  >>> s.value

  4  >>> s.instances

    3

Notes on Classes

  Data hiding is by convention

  Namespaces are inspectable



  >>> dir(sdie.SettableDie)

  [’__doc__’, ’__init__’, ’__module__’, ’getValue’,

’instances’, ’roll’, ’setValue’]

  >>> dir(s)

  [’__doc__’, ’__init__’, ’__module__’, ’getValue’,

  ’instances’, ’roll’, ’setValue’, ’sides’, ’value’]

  Attributes starting with __ are "mangled"

  Attributes starting and ending with __ are special hooks

Documentation Strings (Docstrings)

  Special attribute __doc__ in modules, classes and  functions

  Python libraries are well documented

  >>> from random import randrange

  >>> print randrange.__doc__

  Choose a random item from range(start, stop[, step]).

  This fixes the problem with randint() which includes the    endpoint; in Python this is usually not what you want.

  Do not supply the ’int’ and ’default’ arguments.

  Used by interactive help utility

  >>> help(randrange)

  $ pydoc random.randrange

  Docstrings are easily embedded into new code

  can provide testing framework

Another Class: Just for Fun

  #file:

  """Implementation of a classic stack data structure: class Stack"""  class Stack:

  "Stack implements a classic stack with lists"

    def __init__(self): = []  def push(self, x): .append(x) def top(self): return [-1]  def pop(self): return ()

Advantages for CS1

  Simple language = More time for concepts

  Safe loop and rich built-ins = Interesting programs early

  Free Language and IDE = Easy for students to acquire

  Dynamic features = Ease of experimentation

  Less code = More programming assignments

Our Approach

  Spiral of imperative and OO concepts (objects ontime?)

  Emphasize:

  Algorithmic thinking

  Universal design/programming patterns (not Python)

  Outline

  Simple numeric processing first

  String processing by analogy to numeric

  Using objects via graphics

  Functions and control structures

  Top-down design

  Classes

  Collections

  OO Design

  Algorithm Design and Recursion

Library

  Homegrown 2D graphics package ()

  Thin wrapper over Python standard GUI package Tkinter

  Why?

  Students LOVE graphics, but it adds complexity

  Our package "hides" the event loop

  Teaches graphics and object concepts

  Natural progression

  Learn by using concrete objects

 Build own widgets

  Implement simple event loop

  from graphics import *  # our custom graphics 

  win = GraphWin("Draw a Triangle")  win.setCoords(0.0, 0.0, 10.0, 10.0)  message = Text(Point(5, 0.5), "Click on three points")  (win)  p1 = win.getMouse()  (win)  p2 = win.getMouse()  (win)  p3 = win.getMouse()  (win)  triangle = Polygon(p1,p2,p3)  triangle.setFill("peachpuff")  triangle.setOutline("cyan")  (win)  message.setText("Click anywhere to quit.")

  win.getMouse()

Triangle Screenshot

Face

  Assignment: Draw something with a face

Blackjack Project

Other Approaches to CS1

  Objects First

  Rich set of readily useable objects

  Multi-Paradigm

  Peter Norvig: ’ a dialect of LISP with "traditional" syntax.’

  Breadth-First

  perfect for first brush of programming

  3D Graphics

  VPython -- visualization for mere mortals

  GUI/Events early

  Tkinter is (arguably) the simplest GUI toolkit going

 What About CS2?

  Currently we use Java in CS2

  Why?

  Want our students to see static typing

  Java is a high-demand language

  Switching languages is good for them

  It works

  Students are better programmers coming in

  The conceptual base is the same

  They find Java annoying, but not difficult

  Python is our pseudo-code

  My experience

  CS2 is at least as smooth as before

Python Resources

Liperi, (Prentice Hall)

 "How to Think Like a Computer Scientist: Learning with Python,"

Downey, Elkner, and Meyers (Green Tea Press)

 "Python Programming: An Introduction to Computer Science," Zelle

  (Franklin, Beedle, and Associates)

  Technical Python Books

  Too many to list, see Python web site and Amazon

  Personal Favorite: "Python in a Nutshell," Alex Martelli (O’Reilly and

  Assoc.)

  Python Web Sites

  -- The site for everything Pythonic  ; -- Searchable database of Python add-ons

Conclusions

 Python Rocks! 

 You’ll Never Go Back


0