Transcript chp02

Chapter 2: Design of Algorithms

Representing algorithms


How to express algorithms the best?
Possible solutions:




Natural language




Natural language
Programming language
Pseudo-Code
+ Known by everyone (that speaks it)
-- Hides the structure of the algorithm
-- Allows ambiguity in expressions
Programming language



+ Language that computer understands (e.g. Fortran, C, Pascal)
+ Very exact and well structured
-- Very restrictive particularly in early design phases
1
Representing Algorithms

Pseudo-Code: golden medium



Bases on a natural language (e.g. English)
Uses “programming-language”-like notation (e.g.
statements, conditions etc.)
Examples




Set x to the value 45
Return
Go to step 100 if condition is satisfied
Types of pseudo-code instructions



Sequential operations
Conditional operations
Iterative operations
2
Representing Algorithms

Sequential operations

Input/Output:


Communication with the outside
Input:



Output





Get the value for “variable1” “variable2” …
Example: Get the value of x y z
Print the value for “variable1” “variable2” …
Example: Print the value of x y z
Print the message ‘message content’
Example: Print the message ‘An error has been detected’
Computation:


Set the value of “variable” to “arithmetic operation”
Example: Set the value of Area to p*r2
3
Representing Algorithms

Conditional operations:


If “condition” is true then
Set of algorithmic operations
Else
Set of algorithmic operations
Example:

If denominator = 0 then Print the message ‘Operation cannot
be done’
Else Set fraction to numerator/denominator
4
Representing Algorithms

Iterative operations

Repeat Step i to j until “condition” becomes true
Step i:
…
Step j:

Repeat until “condition” becomes true
Set of operations
End of loop

While “condition” remains true do
Set of operations
End of the loop
5
Representing Algorithms

Examples of iterative operations

Printing squares of 1 to n:


Get the value of n
While n > 0 do



The same algorithm with a repeat loop:


Get the value of n
If n > 0 then


Print n2
Set n to n-1
Repeat until n < 1 becomes true
2
 Print n
 Set n to n-1
Difference between while and repeat:


While: pre-check (continuation) condition then perform work (# of
times 0, 1, 2, …)
Repeat: perform work then post-check (termination) condition (# of
times 1, 2, …)
6
Representing Algorithms

Are the three types of operations (sequential,
conditional, iterative) sufficient to represent ANY
possible valid algorithm?



YES!
Even the most complex algorithms used in e.g.
international switching systems can be represented
using said operations
Compare: We only need few letters (26 in English) to
write the most marvelous novels.
7
Algorithmic Problem Solving

Sequential Search Algorithm

Problem Statement:



Given a name N and a telephone book including names
N1 to N1000 and phone numbers T1 to T1000
Task: Find the phone number T of the name N
First solution:





1.
2.
…
1001.
1002.
Get N, N1,…,N1000, T1, …T1000
If N = N1 then Print the value T1
If N = N1000 then Print the value T1000
Stop
8
Sequential Search Algorithm

Assessment of the first solution




+: Algorithm prints all phone numbers of the
given name N
--: Too long listing (1001 lines)
--: Slow: it checks always all entries even
after the phone number has been already
found
--: No error message if the name is not in the list
9
Sequential Search Algorithm

Second solution:








1. Get N, N1, …, N1000, T1, …, T1000
2. Set i to 1 and mark the N as not yet found
3. Repeat 4 to 7 until N is found
4.
If N = Ni then
5.
Print the Phone number is Ti
6.
Mark N as found
Else
7.
Add 1 to i
8. Stop
10
Sequential Search Algorithm

Assessment of the second solution




+: Very short listing (only 8 lines)
+: Quick if N is in the list, it does not check all entries in all cases
--: Big problem: Endless loop if N is not in the list
Final solution:









1. Get N, N1, …, N1000, T1, …, T1000
2. Set i to 1 and mark the N as not yet found
3. Repeat 4 to 7 until either N is found or i > 1000
4.
If N = Ni then
5.
Print the Phone number is Ti
6.
Mark N as found
Else
7.
Add 1 to i
8. If N is not marked as found then Print message ‘Name is not in
the directory’
9. Stop
11
Finding the Maximum



Given a list of n numbers
Task: find the largest number in the list
Solution:













1. Get the value n
2. Get the values A1, A2, …, An
3. Set the value of maximum to A1
4. Set the value of location to 1
5. Set the value of i to 2
6. While i <= n do
7.
If Ai > maximum then
8.
Set maximum to Ai
9.
Set location to I
10.
Add 1 to I
End of the loop
11. Print maximum and location
12. Stop
12
Pattern Matching



Given a text T (sequence of letters) and a pattern (e.g. word) p
Find the occurrence of p in T
Example:







Text:
Pattern:
Output:
Text:
Pattern:
Output:
let me know whether or not to go
no
Match starting at position 9
let me know whether or not to go
no
Match starting at position 24
Example: Genome research


Genome: … T C G A T T G T C C C A G T G C A A A C T G C A T …
Probe:
AAA
(a match)
13
Pattern Matching

Solution:




1. Get values n and m the size of the text and the pattern
2. Get T1,…,Tn and P1,…,Pm
3. Set k to 1
4. Repeat until k > n-m+1 (until we treated the entire text)


5. Try to match the pattern P1…Pm at position k
6. If there was a match then




7. Print the value of k the starting location of the match
8. Add 1 to k (slides to the right)
End of loop
9. Stop
14
Pattern Matching



Step 5 ???
5. Try to match the pattern P1…Pm at position k
Situation:
Text: T1 T2 … Tk Tk+1…Tk+m-1 Tk+m …Tn
Pattern:
P 1 P2 … Pm

Algorithm for Step 5


Set i to 1 and mismatch to no
Repeat until (i > m) or mismatch = yes

If Pi not equal Tk+i-1 then


Else


Set mismatch to yes
Set i to i+1
End of loop
15