54 lines
1.5 KiB
Java
54 lines
1.5 KiB
Java
|
package DynamicProgramming;
|
||
|
|
|
||
|
|
|
||
|
|
import java.util.Arrays;
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Recursive Solution for 0-1 knapsack with memoization
|
||
|
|
*/
|
||
|
|
public class KnapsackMemoization {
|
||
|
|
|
||
|
|
private static int[][] t;
|
||
|
|
|
||
|
|
|
||
|
|
// Returns the maximum value that can
|
||
|
|
// be put in a knapsack of capacity W
|
||
|
|
public static int knapsack(int[] wt, int[] value, int W, int n) {
|
||
|
|
if(t[n][W] != -1) {
|
||
|
|
return t[n][W];
|
||
|
|
}
|
||
|
|
if (n == 0 || W == 0) {
|
||
|
|
return 0;
|
||
|
|
}
|
||
|
|
if (wt[n - 1] <= W) {
|
||
|
|
t[n-1][W-wt[n-1]] = knapsack(wt, value, W - wt[n - 1], n - 1);
|
||
|
|
// Include item in the bag. In that case add the value of the item and call for the remaining items
|
||
|
|
int tmp1 = value[n - 1] + t[n-1][W-wt[n-1]];
|
||
|
|
// Don't include the nth item in the bag anl call for remaining item without reducing the weight
|
||
|
|
int tmp2 = knapsack(wt, value, W, n - 1);
|
||
|
|
t[n-1][W] = tmp2;
|
||
|
|
// include the larger one
|
||
|
|
int tmp = tmp1 > tmp2 ? tmp1 : tmp2;
|
||
|
|
t[n][W] = tmp;
|
||
|
|
return tmp;
|
||
|
|
// If Weight for the item is more than the desired weight then don't include it
|
||
|
|
// Call for rest of the n-1 items
|
||
|
|
} else if (wt[n - 1] > W) {
|
||
|
|
t[n][W] = knapsack(wt, value, W, n - 1);
|
||
|
|
return t[n][W];
|
||
|
|
}
|
||
|
|
return -1;
|
||
|
|
}
|
||
|
|
|
||
|
|
// Driver code
|
||
|
|
public static void main(String args[]) {
|
||
|
|
int[] wt = {1, 3, 4, 5};
|
||
|
|
int[] value = {1, 4, 5, 7};
|
||
|
|
int W = 10;
|
||
|
|
t = new int[wt.length+1][W+1];
|
||
|
|
Arrays.stream(t).forEach(a -> Arrays.fill(a, -1));
|
||
|
|
int res = knapsack(wt, value, W, wt.length);
|
||
|
|
System.out.println("Maximum knapsack value " + res);
|
||
|
|
}
|
||
|
|
}
|