Recursion Chapter 7 Chapter 7: Recursion2 Chapter Objectives To understand how to think recursively...
-
date post
21-Dec-2015 -
Category
Documents
-
view
223 -
download
0
Transcript of Recursion Chapter 7 Chapter 7: Recursion2 Chapter Objectives To understand how to think recursively...
Chapter 7: Recursion 2
Chapter Objectives
• To understand how to think recursively• To learn how to trace a recursive method• Show how recursion is used in math formulas• To learn how to write recursive algorithms and methods
for searching arrays• To learn about recursive data structures and recursive
methods for a LinkedList class• To understand how to use recursion to solve the Towers
of Hanoi problem
Chapter 7: Recursion 3
Chapter Objectives (continued)
• To understand how to use recursion to process two-dimensional images
• To learn how to apply backtracking to solve search problems such as finding a path through a maze
Chapter 7: Recursion 4
Recursive Thinking
• Recursion is a problem-solving approach that can be used to generate simple solutions to certain kinds of problems that would be difficult to solve in other ways
• Recursion splits a problem into one or more simpler versions of itself
Chapter 7: Recursion 6
Recursive Algorithm
• A given problem is a candidate for recursive solution if you can define a base case and a recursive case • Base case: There is at least one case, for a small
value of n, that can be solved directly• Recursive case: A problem of a given size n can be
split into one or more smaller versions of the same problem
Chapter 7: Recursion 7
Steps to Design a Recursive Algorithm
• Recognize the base case and provide a solution to it• Devise a strategy to
• Split the problem into smaller versions of itself• Smaller versions must progress toward base case
• Then to solve the original problem…• …combine the solutions of the smaller (or smallest)
problems in such a way as to solve the problem
Chapter 7: Recursion 8
Recursive Algorithm to Search Array
• Given: Array of n elements, sorted in increasing order• Replace problem of searching n elements by problem of
searching n/2 elements• To find target element…
• Look at element in the middle• If middle element equals target, then done• Else if target is less than middle element
• Repeat search on first half
• Else (target is greater than middle element)• Repeat search on second half
Chapter 7: Recursion 9
Recursive Algorithm to Search Array
• Suppose we search for “7” in this array of 25 elements:2,3,5,7,11,13,17,19,23,29,31,37,41,43,47,53,59,61,67,71,73,79,83,89,97
• Middle element is 41• Since 7 is less than 41, repeat search using 1st half:
2,3,5,7,11,13,17,19,23,29,31,37
• Middle element is 13• Since 7 is less then 13, repeat using 1st half:
2,3,5,7,11
• Middle element is 5• Since 7 is greater than 5, repeat using 2nd half
7,11
• And so on…
Chapter 7: Recursion 10
Recursive Algorithm to Search Array
• If array is empty• Return -1
• Else if middle element matches target• Return subscript of element
• Else if target is less than middle element• Recursively search first half of array
• Else• Recursively search second half of array
Chapter 7: Recursion 11
General Recursive algorithm
• If the problem can be solved for current value of n• Solve it (base case)
• Else• Recursively apply the algorithm to one or more
problems with smaller value(s) of n (recursive case)
Chapter 7: Recursion 12
Steps to Design Recursive Algorithm
• In general…• A base case (small n) that can be solved directly• Problem of size n can be split into smaller version(s)
of the same problem (recursive case)• To design a recursive algorithm, we must
• Recognize base case, and provide solution to it• Devise a strategy to split problem into smaller
versions of itself• Combine solutions to smaller problems so that large
problem is solved correctly
Chapter 7: Recursion 13
Recursive String Length Algorithm
• How to find length of a string?• Lots of ways to do this• How about a recursive solution?
• Length of empty string is 0• Length of non-empty string is length of first character
plus length of the “rest of the string”• That is, 1 + length of rest of the string
• For example, length of “abcd” is 1 + length of “bcd”• And so on…
Chapter 7: Recursion 15
Recursive String Length Algorithm
• In Java
public static int length(String str) {if (str == null || str.equals(""))
return 0;else
return 1 + length(str.substring(1));}
Chapter 7: Recursion 16
Run-Time Stack and Activation Frames
• “Activation frame” pushed onto run-time stack• Run-time stack is like “scratch paper”• OS can save information (keep state)
• For later use
Chapter 7: Recursion 18
Proving that a Recursive Method is Correct
• Proof by induction• Prove the theorem is true for the base case• Show that if the theorem is assumed true for n, then
it must be true for n+1• Proof of recursion is similar to induction
• Verify that the base case is solved correctly• Verify that each recursive case progresses towards
the base case• Verify that if all smaller problems are solved correctly,
then the original problem is also solved correctly
Chapter 7: Recursion 19
Prove Recursive String Length is Correct
• Base case?• Empty string is of length 0
• Recursive case makes progress towards base case?• String gets smaller by 1 each time
• Show that if smaller problem solved correctly, then original problem is solved correctly• Smaller problem: length(str.substring(1))• If small problem correct, then length of original string
is correct: 1 + length(str.substring(1))
Chapter 7: Recursion 20
Recursive Math Formulas
• Mathematics has many naturally recursive formulas• Examples include:
• Factorial• Powers• Greatest common divisor
• Note that if recursive problem is too big, a stack overflow error will occur• Space available for run-time stack is limited
Chapter 7: Recursion 21
Factorial
• How do you pronounce “n!” ?• As you know, n! = n (n-1) (n-2) … 1
• And 0! = 1• Recursive definition?• We have: n! = n (n-1)!• Can easily write a recursive method for factorial…
Chapter 7: Recursion 23
Exponentiation
• If n is a non-negative integer, xn is x times itself, n times• And x0 = 1
• For example, 27 = 2 2 2 2 2 2 2 = 128• Recursive definition?• We have: xn = x xn-1
• Can easily write recursive method…
Chapter 7: Recursion 25
Recursion Versus Iteration
• What is iteration?• A loop repetition condition determines whether to
repeat the loop body or exit from the loop• In recursion, the condition tests for a base case • There are similarities between recursion and iteration
• You can always write an iterative solution to a problem that is solvable by recursion
• You are probably more familiar with iteration• So, why bother with recursion?
• Recursive code often simpler than an iterative algorithm and thus easier to write, read, and debug
Chapter 7: Recursion 26
Tail Recursion
• “Tail recursion” or “last-line recursion”• Single recursive call, and it is last line of the method
• All of the examples we have considered so far are examples of tail recursion
• Easy to convert tail recursive method to iterative method• Example on the next slide…
Chapter 7: Recursion 28
Efficiency of Recursion
• Is recursion more or less efficient than iteration? • Recursive methods often slower than iteration: Why?• The overhead for loop repetition is smaller than the
overhead for a method call and return• Recall, run-time stack
• If it is easier to conceptualize an algorithm using recursion, then you should code it as a recursive method• The reduction in efficiency does not outweigh the
advantage of readable code that is easy to debug
Chapter 7: Recursion 29
Fibonacci Numbers
• Fibonacci numbers developed to model rabbit population• Defined as
fib1 = 1, fib2 = 1, fibn = fibn-1 + fibn-2
• The first several Fibonacci numbers are
1,1,2,3,5,8,13,21,34,55,89,144,233,377,… • Easy to write a recursive method for Fibonacci numbers
• See next slide…
Chapter 7: Recursion 31
Inefficient Recursive Fibonacci Method
• Why is the obvious Fibonacci recursion inefficient?• Consider fibonacci(7)
• Compute fibonacci(6) and fibonacci(5)• Then fibonacci(5), fibonacci(4), fibonacci(4)
fibonacci(3) • Then fibonacci(4), fibonacci(3), fibonacci(3),
fibonacci(2), fibonacci(3), fibonacci(2), fibonacci(2), fibonacci(1)
• And so on…• Why is this inefficient?
Chapter 7: Recursion 32
Inefficient Recursive Fibonacci
• How inefficient is this?• Exponential time!!!
• If n = 100, need about 2100 activation frames
Chapter 7: Recursion 33
Efficient Recursive Fibonacci Method
• An O(n) Fibonacci algorithm… • If we know current and previous Fibonacci numbers
• Then next Fibonacci number is current + previous• Method fibo
• 1st argument is current Fibonacci (fibCurrent)• 2nd argument is previous Fibonacci (fibPrevious)• 3rd argument is n
• When n = 1, base case, so return fibCurrent• Otherwise, return
fibo(fibCurrent + fibPrevious, fibCurrent, n - 1)
• Start by computing: fibo(1, 0, n)
Chapter 7: Recursion 36
Dynamic Programming
• We want to find “best” stagecoach route from A to J• Where ”best” == shortest distance
Chapter 7: Recursion 37
Dynamic Programming
• Let F(X) be shortest distance from A to X• Note that, for example,
F(J) = min{F(H) + 3, F(I) + 4}• This provides a recursive method to find F(J)…
Chapter 7: Recursion 40
Dynamic Programming
• F(E) = min{F(B) + 7, F(C) + 3, F(D) + 4} = min{9,7,7} = 7
2
4
3
7
4
8
0
• F(F) = min{F(B) + 4, F(C) + 2, F(D) + 1} = min{6,6,4} = 4
• F(G) = min{F(B) + 6, F(C) + 4, F(D) + 5} = min{8,8,8} = 8
Chapter 7: Recursion 41
Dynamic Programming
• F(H) = min{F(E) + 1, F(F) + 6, F(G) + 3} = min{8,10,11} = 8
2
4
3
4
8
8
7
7
0
• F(I) = min{F(E) + 4, F(F) + 3, F(G) + 3} = min{11,7,11} = 7
Chapter 7: Recursion 42
Dynamic Programming
• F(J) = min{F(H) + 3, F(I) + 4} = min{11,11} = 11
2
4
3
7
4
8
8
7
0 11
Chapter 7: Recursion 43
Dynamic Programming
• The shortest path(s) from A to J have distance 11• In this example, the shortest path is not unique
• How to find best path?• As opposed to shortest distance
2
4
3
7
4
8
8
7
110
Chapter 7: Recursion 44
Recursive Array Search
• Searching an array can be accomplished using recursion• Simplest way to search is a linear search
• Examine one element at a time starting with the first element and ending with the last
• Base case for recursive search is an empty array• Result is negative one
• Another base case would be when the array element being examined matches the target
• Recursive step is to search the rest of the array, excluding the element just examined
Chapter 7: Recursion 47
Design of a Binary Search Algorithm
• Binary search can be performed only on an array that has been sorted• We assume array is in increasing order
• Stop cases• The array is empty• Element being examined matches the target
• Check the middle element for a match with the target• Throw away the half of the array
• Throw away the half that the target cannot be in
Chapter 7: Recursion 52
Efficiency of Binary Search
• At each recursive call we eliminate half the array elements from consideration
• O(log2n)
Chapter 7: Recursion 53
Comparable Interface
• Classes that implement the Comparable interface must define a compareTo method that enables its objects to be compared in a standard way• CompareTo allows you to define ordering of elements
Chapter 7: Recursion 54
Method Arrays.binarySearch
• Java API class Arrays contains a binarySearch method• Can be called with sorted arrays of primitive types or
with sorted arrays of objects• If the objects in the array are not mutually comparable
or if the array is not sorted, the results are undefined• If there are multiple copies of the target value in the
array, there is no guarantee which one will be found• Throws ClassCastException if the target is not
comparable to the array elements
Chapter 7: Recursion 55
Recursive Data Structures
• Computer scientists often encounter data structures that are defined recursively• Trees (Chapter 8) are defined recursively
• Linked list can be described as a recursive data structure• Recursive methods useful for processing recursive data
structures• The first language developed for artificial intelligence
research was a recursive language called LISP• “LISt Processing” language
Chapter 7: Recursion 56
Recursive Definition of a Linked List
• A non-empty linked list is a collection of nodes where each node references another linked list consisting of the nodes that follow it in the list
• The last node references an empty list• A linked list is empty, or it contains a node, called the list
head, that stores data and a reference to a linked list
Chapter 7: Recursion 63
Problem Solving with Recursion
• Will look at two problems• Towers of Hanoi• Counting cells in a blob
Chapter 7: Recursion 64
Towers of Hanoi
• Goal: Move stack of disks to one of the empty pegs• Rules:
• Only the top disk on a peg can be moved• A larger disk cannot be placed on top of a smaller disk
Chapter 7: Recursion 65
• Suppose we can get top 2 disks onto middle peg• Then we have smaller version of original problem
Towers of Hanoi
Chapter 7: Recursion 72
Counting Cells in a Blob
• Consider how we might process an image that is presented as a two-dimensional array of color values
• Information in the image may come from• X-Ray• MRI• Satellite imagery• Etc.
• Goal is to determine the size of any area in the image that is considered abnormal because of its color values
Chapter 7: Recursion 78
Backtracking
• Backtracking is an approach to implementing systematic trial and error in a search for a solution• An example is finding a path through a maze
• If you are attempting to walk through a maze, you will probably walk down a path as far as you can go
• Eventually, you will reach your destination or you won’t be able to go any farther
• If you can’t go any farther, you will need to retrace your steps
• Backtracking is a systematic approach to trying alternative paths and eliminating them if they don’t work
Chapter 7: Recursion 79
Backtracking
• Never try the exact same path more than once, and you will eventually find a solution path if one exists
• Problems that are solved by backtracking can be described as a set of choices made by some method
• Recursion allows us to implement backtracking in a relatively straightforward manner• Each activation frame is used to remember the choice
that was made at that particular decision point• A program that plays chess may involve some kind of
backtracking algorithm
Chapter 7: Recursion 86
Chapter Review
• A recursive method has a standard form• To prove that a recursive algorithm is correct, you must
• Verify that the base case is recognized and solved correctly
• Verify that each recursive case makes progress toward the base case
• Verify that if all smaller problems are solved correctly, then the original problem must also be solved correctly
• The run-time stack uses activation frames to keep track of argument values and return points during recursive method calls
Chapter 7: Recursion 87
Chapter Review
• Mathematical Sequences and formulas that are defined recursively can be implemented naturally as recursive methods
• Recursive data structures are data structures that have a component that is the same data structure
• Towers of Hanoi and counting cells in a blob can both be solved with recursion
• Backtracking is a technique that enables you to write programs that can be used to explore different alternative paths in a search for a solution