1 CSCD 300 Data Structures Recursion. 2 Proof by Induction Introduction only - topic will be covered in detail in CS 320 Prove: N   i = N ( N + 1.

Presentation on theme: "1 CSCD 300 Data Structures Recursion. 2 Proof by Induction Introduction only - topic will be covered in detail in CS 320 Prove: N   i = N ( N + 1."— Presentation transcript:

1 CSCD 300 Data Structures Recursion

2 Proof by Induction Introduction only - topic will be covered in detail in CS 320 Prove: N   i = N ( N + 1 ) / 2 i=1 Basis Step : – show true for trivial case - here N = 1 for N = 1 the sum is 1( 1 + 1 ) / 2 = 1

3 Proof by Induction Inductive hypothesis : – statement is true for N = k Inductive step : if true for N = k show true for N = k + 1 – k+1 k   i = (k + 1) +  i i=1 i=1

4 Proof by Induction Inductive step : k+1   i = (k + 1) + k(k + 1) / 2 = (k + 1) (k + 2) /2 i=1 – so by induction the statement is true for all k

5 Recursion Math definition: – a solution to a problem that is defined in terms of a simpler version of itself Programming definition: – a function which calls itself

6 Recursion Recursive definition of Sum of Integers: public static int Sum ( int n ) { // assumes n is >= 1 if( n == 1) return 1; else return Sum( n - 1 ) + n; }

7 Factorial : Mathematical Definition Mathematical definition (recursive): 1 if N=1 or N=0 N! = N * (N - 1)! if N > 1

8 Factorial: Recursive Method Recursive method : public static int factorial ( int n ) { int temp; System.out.println("Entering factorial: n = " + n); if((n == 1)|| (n == 0)) temp = 1; else temp = n * factorial(n-1); System.out.println("Leaving factorial: n = "+ n ); return temp; } Call: System.out.println("Factorial(4):" + factorial(4));

9 Stack Frames System Stack – used to control which function is currently active and to reverse the order of function calls when returning. Stack Frame – a variable size piece of memory that is pushed onto the system stack each time a function call is made.

10 Stack Frames Stack Frame contains: space for all local (automatic) variables the return address - the place in the program to which execution will return when this function ends the return value from the function all parameters for a function (with actual parameter values copied in – The stack frame on top of the stack always represents the function being executed at any point in time - only the stack frame on top of the stack is accessible at any time.

11 Factorial Trace Entering factorial N = 4 Entering factorial N = 3 Entering factorial N = 2 Entering factorial N = 1 Leaving factorial N = 1 Leaving factorial N = 2 Leaving factorial N = 3 Leaving factorial N = 4 Ret Addr N = 1 Ret Value : 2 1 2 Ret Addr N = 2 2 Ret Value : Ret Addr N = 3 2 Ret Value : Ret Addr N = 4 1 Ret Value : 2 6 24

12 Rules of Recursion Always include a terminal case which does not require recursion to calculate Each recursive call should progress toward the terminal case Assume each recursive call works when designing a recursive algorithm

13 Recursive Array Printing public class ArrayPrintDriver { public static void printem( int[ ] array, int n, int last ) { if(n <= last){ System.out.println( array[n]); printem(array, n + 1, last); } public static void main( String args[] ) { int[] values = new int[5]; for(int i = 0; i < 5 ; i++) values[i] = i+1; printem(values, 0, 4); } 2 1

14 Array Printing Trace e 1 2 3 4 5 Ret Addr last = 4 n = 0 1 array = values Ret Addr last = 4 n = 1 2 array = values Ret Addr last = 4 n = 2 2 array = values Ret Addr last = 4 n = 3 2 array = values Ret Addr last = 4 n = 4 2 array = values

15 Multiple Recursion Fibonacci Sequence: 0 1 1 2 3 5 8 13 21 44 – each number in sequence is the sum of the previous two numbers (except the first two) – Fibonacci number 0 = 0 Fibonacci number 1 = 1 – all other Fibonacci numbers are defined as the sum of the previous two

16 Multiple Recursion public static int fib ( int n ) { if ( n <= 1) return n; else return Fib( n - 1) + Fib( n - 2 ); } Problem: multiple calls are made to calculate the same Fibonacci number

17 Problems with Recursion Memory (stack) exhaustion: – if many stack frames are pushed - can run out of space for stack – if each call allocates much memory - either automatically or dynamically - heap exhaustion is possible Time: – each recursive call requires time to set up a new stack frame, copy in actual parameter values and transfer execution

18 Problems with Recursion Any recursive algorithm can be rewritten using iterative technique - it will be longer and more likely to have problems but will run faster.

19 Recursive Binary Search public static int binarySearch(Comparable[ ] anArray, Comparable target, int left, int right) { if(right < left) return -1; // target not present else { int mid = (left + right) / 2; if( target.compareTo(anArray[mid]) == 0) return mid; if( target.compareTo(anArray[mid]) < 0) return binarySearch(anArray, target, left, mid -1); else return binarySearch(anArray, target, mid+1, right); } // end else }

20 Binary Search Trace 0235 7 910131520 [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Call 1 - left = 0 right = 9 mid = 4 target = 3 0 2 35791013 15 20 [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Call 2 - left = 0 right = 3 mid = 1 02 3 57910 13 15 20 [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Call 3 - left = 2 right = 3 mid = 2 return 2 (successful search)

21 Binary Search Trace 2 02 35 7 91013 15 20 [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Call 1 - left = 0 right = 9 mid = 4 target = 1 0235 79 10131520 [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Call 2 - left = 0 right = 3 mid = 1 0 2 3 5 79 101315 20 [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Call 3 - left = 0 right = 0 mid = 0 Call 4 - left = 1 right = 0 Return -1 -- unsuccessful search

22 Binary Search Analysis Worst case -target is not found: – Analysis proceeds in the same way as for the iterative version only the number of passes is replaced by the number of calls to binarySearch.

23 Towers of Hanoi Problem - disks of varying diameter are all on one peg (of three) with the largest disks on the bottom and the smallest on the top Goal - move all disks from peg A to peg C Rules - must move one disk at a time - can not place a larger disk on a smaller disk ABC 1 ABC 2 ABC 3

24 Towers of Hanoi - Stages ABC 4 ABC 5 ABC 6 ABC 7 ABC 8

25 Towers of Hanoi - Algorithm public static void hanoi(int n, char source, char dest, char intermed) { if (n == 1) System.out.println("move a disk from " + source + " to " + dest); else { hanoi(n-1, source, intermed, dest); System.out.println("move a disk from " + source + " to " + dest); hanoi(n-1, intermed, dest, source); } original call: hanoi ( 3, 'A', 'C', 'B'); 1 2 3

26 Towers of Hanoi - Analysis How many calls are made to hanoi for n=3 ? – Each non-terminal call generates two more. – 2 n - 1 calls generated so f(n) = 2 n - 1 – Algorithm has O(2 n ) time complexity (exponential) n = 3 n = 2 n = 1

Download ppt "1 CSCD 300 Data Structures Recursion. 2 Proof by Induction Introduction only - topic will be covered in detail in CS 320 Prove: N   i = N ( N + 1."

Similar presentations