While waiting: quotes

1
While waiting quotes

• 640K of main memory ought to be enough for
  anybody. - W. Gates (Founder and CEO Microsoft),
  1981 - disclaimed
• We flew down weekly to meet with IBM, but they
  thought the way to measure software was the
  amount of code we wrote, when really the better
  the software, the fewer lines of code. - W. Gates
• Beauty is more important in computing than
  anywhere else in technology because software is
  so complicated. Beauty is the ultimate defense
  against complexity. - D. Gelernter ("Machine
  Beauty", Basic Books, 1998)
• Good programmers know what's beautiful and bad
  ones don't. - D. Gelernter ("Machine Beauty",
  Basic Books, 1998)
• Vague and nebulous is the beginning of all
  things, but not their end. - K. Gibran
• The purpose of computing is insight, not numbers.
  - R. Hamming

2
COS 121 - Nov. 09, 2005

• A file that big?
• It might be very useful.
• But now it is gone.

Stefan Brandle
3
Agenda

• Read Ch. 9.2 (pp. 465-490) for Friday mergesort,
  quicksort, radix sort, algorithm comparison
• Guest speaker for WOW
• MP 3 -- N-queens speed trial
• Searching linear, binary
• Sorting Selection, Bubble, Insertion

4
Guest Speaker

• A guest who builds tents via commercial ventures
• Keeping name out of lecture notes, because they
  get posted online.
• This is an important thing to think about.

5
MP 3 Questions?

• Be sure to read through the Wirth Program
  Development by Stepwise Refinement paper
• Any questions?
• Due date is Friday, midnight.

6
Searching

• Linear
• Best case 1(not particularly likely or
  meaningful)
• Average case O(1/2 N) O(N) Drop coefficients
  of N when computing order of magnitude
• Worst case O(N)
• Binary
• Suppose that N 2k for some k. E.g., 8 23
• Given N, can divide N by 2, k times, before get
  down to 1 item. So k log2N
• So binary search is, worst case, O(log2N)
• If N is not a power of 2, then find 2k-1 lt N lt
  2kE.g., for N7, k3.

7
Selection Sort

• Each pass through the unsorted data, select the
  largest item. Swap it into the sorted area.
• Demonstrate the sort with plastic cups.
• Look at code pp. 458-459.
• Analysis
• Various ways of measuring algorithmic work, but a
  good approximation is to count number of
  comparisons and assignments
• Calls to indexOfLargest (n-1) (n-2) 1
  n (n - 1)/2
• Each of n-1 calls to swap require 3
  assignments/data moves.
• gt n(n-1)/2 3(n-1) n2/2 5n/2 - 3 gt
  O(n2)

8
Bubble Sort

• Similar to selection sort. Start at one end of
  unsorted data and start bubbling an item up to
  top of data. Compare two items, swap if
  necessary, so that smaller is closer to top. Then
  compare the smaller of the two with the next
  position and swap if necessary. Keep doing until
  smallest item in array is at top (front) of
  array.
• Demonstrate with cups.
• Look at code p. 462.
• Analysis pass i requires n-i comparisons and at
  most n-i swaps.gt n (n-1)/2 comparisons, same
  number of swapsgt 2 n (n-1) 2 n2 - 2n gt
  O(n2)

9
Insertion Sort

• Take a data item from the unsorted part of the
  array. Insert it into the sorted part of the
  array. Repeat until all unsorted data has been
  placed correctly.
• Run through with plastic cups.
• Look at code on pp. 464-465.
• Analysis n(n-1)/2 comparisons 2(n-1)
  moves n2 n - 2gt O(n2)