Develop Reference

Generate all possible combinations longer than the length of an array

I need to build each combination of length L from an String Array/ArrayList, where L is greater than the Array length
I currently have a recursive method (not of my own creation) that will generate each combination of a String[], as long as the combinations are shorter than the Array.
example/psudoCode:
input (2, {A,B,C})
returns {AA, AB, AC, BA, BC, CB, CA}
As of now, if the requested combination length (2 in the example) is greater than the Array length (4,5,6... instead of 2), the recursive method shoots out that sweet sweet ArrayIndexOutOfBounds error.
What I need is a method (recursive or not) that will return every combination of the array, regardless of whether the combinations are longer than the Array itself. Would this be done better by adding more letters to the Array and crossing my fingers or is there a legitimate way to accomplish this? Thank you!
Here is the method I have been using. If u know where the credit lies please say so, this is not of my own creation.
public class bizzBam
{
// Driver method to test below methods
public static void main(String[] args) {
System.out.println("First Test");
String set1[] = {"a", "b","c"};
printAllKLength(set1, pointX);
}
// The method that prints all possible strings of length k. It is
// mainly a wrapper over recursive function printAllKLengthRec()
static void printAllKLength(String set[], int k) {
int n = set.length+2;
printAllKLengthRec(set, "", n, k);
}
// The main recursive method to print all possible strings of length k
static void printAllKLengthRec(String set[], String prefix, int n, int length) {
// Base case: k is 0, print prefix
if (length == 0) {
System.out.println(prefix);
return;
}
// One by one add all characters from set and recursively
// call for k equals to k-1
for (int i = 0; i < n; ++i) {
// Next character of input added
String newPrefix = prefix + set[i];
// k is decreased, because we have added a new character
printAllKLengthRec(set, newPrefix, n, length - 1);
}
}
}
(Edit forgot to say:)
For this algorithim at least, if "PointX" is greater than the input array's length, it will return the indexoutofbounds.

Strictly speaking these are permutations rather than combinations. You're generating all permutations of k elements selected from a set of n candidates, with replacement (or repitition). There will be n^k such permutations.
Here's a non-recursive solution.
public class Permutations
{
public static void main(String[] args)
{
permutationsKN(new String[]{"a", "b", "c"}, 4);
}
static void permutationsKN(String[] arr, int k)
{
int n = arr.length;
int[] idx = new int[k];
String[] perm = new String[k];
while (true)
{
for(int i=0; i<k; i++) perm[i] = arr[idx[i]];
System.out.println(String.join("", perm));
// generate the next permutation
int i = idx.length - 1;
for (; i >= 0; i--)
{
idx[i]++;
if (idx[i] < n) break;
idx[i] = 0;
}
// if the first index wrapped around then we're done
if (i < 0) break;
}
}
}

You have two problems here:
int n = set.length+2; -> This is giving you your "sweet sweet" IndexArrayOutOfBoundsException. Change it to set.length-1. I am not sure why you decided to randomnly put +2 there.
for (int i = 0; i < n; ++i) -> You will be looping from 0 to n. You need to loop from 0 to n-1.
Edit: Or as #SirRaffleBuffle suggested, just do set.length. Total credits to him

Assuming your example is missing "BB" and "CC" because it includes "AA", it looks like what you want is just like the odometer of a car except that instead of ten digits, you want a choice of letters. It's not hard to model an odometer:
class Odo {
private final char [] chars;
private final int [] positions;
private boolean hasNext;
Oddo(String chars, int nPositions) {
this.chars = chars.toCharArray();
this.positions = new int [nPositions];
this.hasNext = true;
}
boolean hasNext() {
return hasNext;
}
String emitNext() {
StringBuilder sb = new StringBuilder();
for (int i = 0; i < positions.length; ++i) sb.append(chars[positions[i]]);
for (int i = 0; i < positions.length; ++i) {
if (++positions[i] < chars.length) {
hasNext = true;
return sb.toString();
}
positions[i] = 0;
}
hasNext = false;
return sb.toString();
}
}
Calling like so:
Odo odo = new Odo("AB", 3);
while (odo.hasNext()) {
System.out.println(odo.emitNext());
}
Produces
AAA
BAA
ABA
BBA
AAB
BAB
ABB
BBB

O(log n) Programming

I am trying to prepare for a contest but my program speed is always dreadfully slow as I use O(n). First of all, I don't even know how to make it O(log n), or I've never heard about this paradigm. Where can I learn about this?
For example,
If you had an integer array with zeroes and ones, such as [ 0, 0, 0, 1, 0, 1 ], and now you wanted to replace every 0 with 1 only if one of it's neighbors has the value of 1, what is the most efficient way to go about doing if this must occur t number of times? (The program must do this for a number of t times)
EDIT:
Here's my inefficient solution:
import java.util.Scanner;
public class Main {
static Scanner input = new Scanner(System.in);
public static void main(String[] args) {
int n;
long t;
n = input.nextInt();
t = input.nextLong();
input.nextLine();
int[] units = new int[n + 2];
String inputted = input.nextLine();
input.close();
for(int i = 1; i <= n; i++) {
units[i] = Integer.parseInt((""+inputted.charAt(i - 1)));
}
int[] original;
for(int j = 0; j <= t -1; j++) {
units[0] = units[n];
units[n + 1] = units[1];
original = units.clone();
for(int i = 1; i <= n; i++) {
if(((original[i - 1] == 0) && (original[i + 1] == 1)) || ((original[i - 1] == 1) && (original[i + 1] == 0))) {
units[i] = 1;
} else {
units[i] = 0;
}
}
}
for(int i = 1; i <= n; i++) {
System.out.print(units[i]);
}
}
}

This is an elementary cellular automaton. Such a dynamical system has properties that you can use for your advantages. In your case, for example, you can set to value 1 every cell at distance at most t from any initial value 1 (cone of light property). Then you may do something like:
get a 1 in the original sequence, say it is located at position p.
set to 1 every position from p-t to p+t.
You may then take as your advantage in the next step that you've already set position p-t to p+t... This can let you compute the final step t without computing intermediary steps (good factor of acceleration isn't it?).
You can also use some tricks as HashLife, see 1.

As I was saying in the comments, I'm fairly sure you can keep out the array and clone operations.
You can modify a StringBuilder in-place, so no need to convert back and forth between int[] and String.
For example, (note: This is on the order of an O(n) operation for all T <= N)
public static void main(String[] args) {
System.out.println(conway1d("0000001", 7, 1));
System.out.println(conway1d("01011", 5, 3));
}
private static String conway1d(CharSequence input, int N, long T) {
System.out.println("Generation 0: " + input);
StringBuilder sb = new StringBuilder(input); // Will update this for all generations
StringBuilder copy = new StringBuilder(); // store a copy to reference current generation
for (int gen = 1; gen <= T; gen++) {
// Copy over next generation string
copy.setLength(0);
copy.append(input);
for (int i = 0; i < N; i++) {
conwayUpdate(sb, copy, i, N);
}
input = sb.toString(); // next generation string
System.out.printf("Generation %d: %s\n", gen, input);
}
return input.toString();
}
private static void conwayUpdate(StringBuilder nextGen, final StringBuilder currentGen, int charPos, int N) {
int prev = (N + (charPos - 1)) % N;
int next = (charPos + 1) % N;
// **Exactly one** adjacent '1'
boolean adjacent = currentGen.charAt(prev) == '1' ^ currentGen.charAt(next) == '1';
nextGen.setCharAt(charPos, adjacent ? '1' : '0'); // set cell as alive or dead
}
For the two samples in the problem you posted in the comments, this code generates this output.
Generation 0: 0000001
Generation 1: 1000010
1000010
Generation 0: 01011
Generation 1: 00011
Generation 2: 10111
Generation 3: 10100
10100

The BigO notation is a simplification to understand the complexity of the Algorithm. Basically, two algorithms O(n) can have very different execution times. Why? Let's unroll your example:
You have two nested loops. The outer loop will run t times.
The inner loop will run n times
For each time the loop executes, it will take a constant k time.
So, in essence your algorithm is O(k * t * n). If t is in the same order of magnitude of n, then you can consider the complexity as O(k * n^2).
There is two approaches to optimize this algorithm:
Reduce the constant time k. For example, do not clone the whole array on each loop, because it is very time consuming (clone needs to do a full array loop to clone).
The second optimization in this case is to use Dynamic Programing (https://en.wikipedia.org/wiki/Dynamic_programming) that can cache information between two loops and optimize the execution, that can lower k or even lower the complexity from O(nˆ2) to O(n * log n).

How can I count if a specific set of digits are present in an array

Suppose I have a number 123. I need to see if I get all digits 1 through 9, including 0. The number 123 has three digits: 1,2, and 3. Then I multiply it by 2 and get 246 (I get digits 2, 4, 6). Then I multiply it by 3 and I get 369. I keep doing incremental multiplication until I get all digits.
My approach is the following:
public int digitProcessSystem(int N) {
String number = Integer.toString(N);
String [] arr = number.split("");
// List <Integer> arr2 = new ArrayList<>();
for (Integer i = 0; i < arr.length; i++) {
try {
arr2[i] = Integer.parseInt(arr[i]);
} catch (NumberFormatException e) {
}
}
count =0;
boolean contains = IntStream.of(arr2).anyMatch(x -> x == 1|| x==2 ||x == 3|| x==4|| x == 5|| x==6 ||x == 7|| x==8||x == 9|| x==0);
}
I really don't know how can I keep doing the boolean for digits that did not match in the first trail above because I will definitely get any one of the all digits in the above boolean search. How can I get that if some specific digits are present and some are not so that I can multiply the actual number to do the search for the digits that were not found in the first trial; just like the way I defined in the beginning.

You could wrap that into a while loop and include the numbers into a Set. Once the set has the size 10 all digits are present in the number. I´d also suggest to use a long instead of an int or you´ll be getting wrong results or run into an excpetion. Here´s some example code for this:
private static long digitProcessSystem(long N) {
long numberN = N;
String number = Long.toString(N);
// calculate 10 digits number here yet
if (number.length() < 10) {
// using the smallest possible number with each digit
// By using this number we are most likely allmost at the result
// This will increase the performance for small digits heavily.
long divider = 1023456789L / numberN;
numberN *= divider;
}
number = Long.toString(numberN);
String[] arr = number.split("");
Set<String> input = new HashSet<>(Arrays.asList(arr));
while(input.size() != 10){
// add N to number
numberN += N;
// Parse the new number
number = Long.toString(numberN);
// split
arr = number.split("");
// clear set
input.clear();
// Add the new numbers to the set. If it has the size 10 now the loop will stop and return the number.
input.addAll(Arrays.asList(arr));
};
return numberN;
}
public static void main(String[] args) {
System.out.println(digitProcessSystem(123));
}
output:
1023458769

I'm not sure what is your end goal. But you can use a HashSet and do something like this in order to achieve what you are trying to achieve:
public static void main (String[] args) throws Exception {
long number = 123L, counter = 1000000000L / number;
while(digitProcessSystem(number * counter++));
System.out.println("Number: " + number * (counter - 1));
}
public static boolean digitProcessSystem(long input) {
char[] arr = Long.toString(input).toCharArray();
Set<Character> set = new HashSet<>();
for (int i = 0; i < arr.length; i++) {
set.add(arr[i]);
}
return set.size() != 10;
}
Output:
Number: 1023458769

without using java language Facilities and hashset:
private static long digitProcessSystem(long N) {
long numberN = N;
String number = Long.toString(N);
String[] arr = number.split("");;
int arr2=new int[10];
int sum=0;
while(sum != 10){
sum=0;
// add N to number
numberN += N;
// Parse the new number
number = Long.toString(numberN);
// If it doesn´t have 10 digitis continue here yet
if(number.length() < 10) continue;
// split
arr = number.split("");
for(int i=0;i<arr.length;i++){
arr2[arr]=1;
}
for(int i=0;i<10;i++){
sum+=arr2[i];
}
};
return numberN;
}

Generally, if you want to process the characters of a String, don’t do it by splitting the string into substrings. Note that every CharSequence, including String, has the methods chars() and codepoints() allowing to process all characters as IntStream.
To check whether all digits from '0' to '9' are present, we can use chars() (don’t have to think about surrogate pairs) and do it straight-forward, map them to their actual number by subtracting '0', filter out all non-digits (just to be sure), then, map them to an int where the nth bit is set, so we can binary or them all together and check whether all of the lowest ten bits are set:
public static boolean hasAllDigits(String s) {
return s.length()>9 &&
s.chars().map(c -> c-'0').filter(c -> c>=0 && c<=9)
.map(c -> 1 << c).reduce(0, (a,b)->a|b) == 0b1111111111;
}
As a bonus, a length-check is prepended as a String must have at least ten characters to contain all ten digits, so we can short-cut if it hasn’t.
Now, I’m not sure about your actual task. If you just want to iterate until encountering a number having all digits, it’s quite simple:
long number=123;
for(long l = 1, end = Long.MAX_VALUE/number; l < end; l++) {
long candidate = number * l;
if(hasAllDigits(String.valueOf(candidate))) {
System.out.println("found: "+candidate);
return;
}
}
System.out.println("not found within the long range");
But if you want to know when you encountered all digits within the sequence of numbers, we have to adapt the test method and keep the bitset between the iterations:
public static int getDigits(String s) {
return s.chars().map(c -> c-'0').filter(c -> c>=0 && c<=9)
.map(c -> 1 << c).reduce(0, (a,b)->a|b);
}
 
long number=123;
int digits=0;
for(long l = 1, end = Long.MAX_VALUE/number; l < end; l++) {
long candidate=number * l;
int newDigits=digits | getDigits(String.valueOf(candidate));
if(newDigits != digits) {
System.out.printf("pos %10d: %10d%n", l, candidate);
digits=newDigits;
if(digits == 0b1111111111) {
System.out.println("encountered all digits");
break;
}
}
}
if(digits != 0b1111111111) {
System.out.println("did not encounter all digits within the long range");
}
This method will only print numbers of the sequence which have at least one digit not encountered before, so you can easily see which one contributed to the complete set and will see at most ten numbers of the sequence.

Generate all possible string from a given length

I would like to be able to generate all possible strings from a given length, and I frankly don't know how to code that. So for further explanation, I and a friend would like to demonstrate some basic hacking techniques, so bruteforcing comes up. Of course, he will be my victim, no illegal thing there.
However, the only thing he told me is that his PW will be 4-char-long, but I'm pretty sure his PW won't be in any dictionnary, that would be toi easy.
So I came up with the idea of generating EVERY 4-char-long-string possible, containing a-z characters (no caps).
Would someone have a lead to follow to code such an algorithm ? I don't really bother with performances, if it takes 1 night to generate all PW, that's no problem.
Don't forget, that's only on demonstration purposes.

You can do it just how you'd do it with numbers. Start with aaaa. Then increment the 'least significant' part, so aaab. Keep going until you get to aaaz. Then increment to aaba. Repeat until you get to zzzz.
So all you need to do is implement is
String getNext(String current)
To expand on this; It possibly isnt the quickest way of doing things, but it is the simplest to get right.
As the old adage goes - 'first make it right, then make it fast'. Getting a working implementation that passes all your tests (you do have tests, right?) is what you do first. You then rewrite it to make it fast, using your tests as reassurance you're not breaking the core functionality.

The absolutely simplest way is to use four nested loops:
char[] pw = new char[4];
for (pw[0] = 'a' ; pw[0] <= 'z' ; pw[0]++)
for (pw[1] = 'a' ; pw[1] <= 'z' ; pw[1]++)
for (pw[2] = 'a' ; pw[2] <= 'z' ; pw[2]++)
for (pw[3] = 'a' ; pw[3] <= 'z' ; pw[3]++)
System.out.println(new String(pw));
This does not scale well, because adding extra characters requires adding a level of nesting. Recursive approach is more flexible, but it is harder to understand:
void findPwd(char[] pw, int pos) {
if (pos < 0) {
System.out.println(new String(pwd));
return;
}
for (pw[pos] = 'a' ; pw[pos] <= 'z' ; pw[pos]++)
findPwd(pw, pos-1);
}
Call recursive method like this:
char[] pw = new char[4];
findPwd(pw, 3);

private static void printAllStringsOfLength(int len) {
char[] guess = new char[len];
Arrays.fill(guess, 'a');
do {
System.out.println("Current guess: " + new String(guess));
int incrementIndex = guess.length - 1;
while (incrementIndex >= 0) {
guess[incrementIndex]++;
if (guess[incrementIndex] > 'z') {
if (incrementIndex > 0) {
guess[incrementIndex] = 'a';
}
incrementIndex--;
}
else {
break;
}
}
} while (guess[0] <= 'z');
}

public class GenerateCombinations {
public static void main(String[] args) {
List<Character> characters = new ArrayList<Character>();
for (char c = 'a'; c <= 'z'; c++) {
characters.add(c);
}
List<String> allStrings = new ArrayList<String>();
for (Character c : characters) {
for (Character d : characters) {
for (Character e : characters) {
for (Character f : characters) {
String s = "" + c + d + e + f;
allStrings.add(s);
}
}
}
}
System.out.println(allStrings.size()); // 456 976 combinations
}
}

This is something you can do recursively.
Lets define every (n)-character password the set of all (n-1)-character passwords, prefixed with each of the letters a thru z. So there are 26 times as many (n)-character passwords as there are (n-1)-character passwords. Keep in mind that this is for passwords consisting of lower-case letters. Obviously, you can increase the range of each letter quite easily.
Now that you've defined the recursive relationship, you just need the terminating condition.
That would be the fact that there is only one (0)-character password, that being the empty string.
So here's the recursive function:
def printNCharacterPasswords (prefix, num):
if num == 0:
print prefix
return
foreach letter in 'a'..'z':
printNCharacterPasswords (prefix + letter, num - 1)
to be called with:
printNCharacterPasswords ("", 4)
And, since Python is such a wonderful pseudo-code language, you can see it in action with only the first five letters:
def printNCharacterPasswords (prefix, num):
if num == 0:
print prefix
return
for letter in ('a', 'b', 'c', 'd', 'e'):
printNCharacterPasswords (prefix + letter, num - 1)
printNCharacterPasswords ("", 2)
which outputs:
aa
ab
ac
ad
ae
ba
bb
bc
bd
be
ca
cb
cc
cd
ce
da
db
dc
dd
de
ea
eb
ec
ed
ee

A aroth points out, using a digit counter approach is faster. To make this even faster, you can use a combination of an inner loop for the last digit and a counter for the rest (so the number of digits can be variable)
public static void main(String... args) {
long start = System.nanoTime();
int letters = 26;
int count = 6;
final int combinations = (int) Math.pow(letters, count);
char[] chars = new char[count];
Arrays.fill(chars, 'a');
final int last = count - 1;
OUTER:
while (true) {
for (chars[last] = 'a'; chars[last] <= 'z'; chars[last]+=2) {
newComination(chars);
chars[last]++;
newComination(chars);
}
UPDATED:
{
for (int i = last - 1; i >= 0; i--) {
if (chars[i]++ >= 'z')
chars[i] = 'a';
else
break UPDATED;
}
// overflow;
break OUTER;
}
}
long time = System.nanoTime() - start;
System.out.printf("Took %.3f seconds to generate %,d combinations%n", time / 1e9, combinations);
}
private static void newComination(char[] chars) {
}
prints
Took 0.115 seconds to generate 308,915,776 combinations
Note: the loop is so simple, its highly likely that the JIT can eliminate key pieces of code (after in-lining newCombination) and the reason its so fast is its not really calculating every combination.
A simpler way to generate combinations.
long start = System.nanoTime();
int letters = 26;
int count = 6;
final int combinations = (int) Math.pow(letters, count);
StringBuilder sb = new StringBuilder(count);
for (int i = 0; i < combinations; i++) {
sb.setLength(0);
for (int j = 0, i2 = i; j < count; j++, i2 /= letters)
sb.insert(0, (char) ('a' + i2 % letters));
// System.out.println(sb);
}
long time = System.nanoTime() - start;
System.out.printf("Took %.3f seconds to generate %,d combinations%n", time / 1e9, combinations);
prints
aaaa
aaab
aaac
....
zzzx
zzzy
zzzz
Took 0.785 seconds to generate 456,976 combinations
It spends most of its time waiting for the screen to update. ;)
If you comment out the line which prints the combinations, and increase the count to 5 and 6
Took 0.671 seconds to generate 11,881,376 combinations
Took 15.653 seconds to generate 308,915,776 combinations

public class AnagramEngine {
private static int[] anagramIndex;
public AnagramEngine(String str) {
AnagramEngine.generate(str);
}
private static void generate(String str) {
java.util.Map<Integer, Character> anagram = new java.util.HashMap<Integer, Character>();
for(int i = 0; i < str.length(); i++) {
anagram.put((i+1), str.charAt(i));
}
anagramIndex = new int[size(str.length())];
StringBuffer rev = new StringBuffer(AnagramEngine.start(str)+"").reverse();
int end = Integer.parseInt(rev.toString());
for(int i = AnagramEngine.start(str), index = 0; i <= end; i++){
if(AnagramEngine.isOrder(i))
anagramIndex[index++] = i;
}
for(int i = 0; i < anagramIndex.length; i++) {
StringBuffer toGet = new StringBuffer(anagramIndex[i] + "");
for(int j = 0; j < str.length(); j++) {
System.out.print(anagram.get(Integer.parseInt(Character.toString(toGet.charAt(j)))));
}
System.out.print("\n");
}
System.out.print(size(str.length()) + " iterations");
}
private static boolean isOrder(int num) {
java.util.Vector<Integer> list = new java.util.Vector<Integer>();
String str = Integer.toString(num);
char[] digits = str.toCharArray();
for(char vecDigits : digits)
list.add(Integer.parseInt(Character.toString(vecDigits)));
int[] nums = new int[str.length()];
for(int i = 0; i < nums.length; i++)
nums[i] = i+1;
for(int i = 0; i < nums.length; i++) {
if(!list.contains(nums[i]))
return false;
}
return true;
}
private static int start(String str) {
StringBuffer num = new StringBuffer("");
for(int i = 1; i <= str.length(); i++)
num.append(Integer.toString(i));
return Integer.parseInt(num.toString());
}
private static int size(int num) {
int size;
if(num == 1) {
return 1;
}
else {
size = num * size(num - 1);
}
return size;
}
public static void main(final String[] args) {
final java.util.Scanner sc = new java.util.Scanner(System.in);
System.out.printf("\n%s\t", "Entered word:");
String word = sc.nextLine();
System.out.printf("\n");
new AnagramEngine(word);
}
}

Put all the characters you expect the password to contain into an array. Write a stub function to test if your algorithm finds the correct password. Start with passwords of length 1, work your way up to 4 and see if your fake password is found on each iteration.

you can use the following code for getting random string. It will return you a string of 32 chars. you can get string of desired length by using substring(). Like if you want a string with 10 chars then:
import java.security.SecureRandom;
import java.math.BigInteger;
SecureRandom srandom = new SecureRandom();
String rand = new BigInteger(176, srandom).toString(32);
rand.substring(0,7);

uppercase random number of characters from a random alphanumeric string

I have some random string with unknown content, what is known is that the content is alphanumeric and in lower case.
I am looking for a simple method to upper case a random number of the alpha characters in that string. The higher the randomness the better.
I can think of a few ways to do this, but none of them seem very optimal.
alright first solution:
public String randomizeCase(String myString){
Random rand = new Random();
StringBuilder build = new StringBuilder();
for(char c: myString.toCharArray()){
String s = new String(c);
if(Character.isLetter(c) && rand.nextBoolean()){
s = s.toUpperCase();
}
build.append(s);
}
return build.toString();
}
I dont like this solution because:
50% chance that every char is uppercased does not equal 50% chance that 50% of the chars are uppercased
There is a chance that nothing is upped cased
char to string conversion is ugly

The solution depends on the probabilistic model you choose. If for example you decide on binomial distribution, then you can traverse the chars, and switch every char to uppercase with a fixed probability p. The expected number of uppercase letters will be p * str.length():
public static String randomUpper(String str, double p) {
StringBuilder sb = new StringBuilder(str.length());
for (int i = 0; i < str.length(); i++) {
char c = str.charAt(i);
if (Character.isLetter(c) && Math.random() < p)
c = Character.toUpperCase(c);
sb.append(c);
}
return sb.toString();
}
If on the other hand you want to decide on the exact number of upercase letters for a given string, then the problem becomes a random sample problem (i.e. choose M positions to switch out of N positions in the string). This can be much faster than the first approach, when M is much smaller than N (though with Java's immutable strings the difference becomes minor because you have to copy the whole string anyway).
-- edit --
Now that you clarified the requirements, consider the following:
public static String randomUpper2(String str, double p) {
int letters = 0;
for (int i = 0; i < str.length(); i++) {
if (Character.isLetter(str.charAt(i)))
letters++;
}
int toChoose = (int) (p * letters);
StringBuilder sb = new StringBuilder(str.length());
for (int i = 0; i < str.length(); i++) {
char c = str.charAt(i);
if (Character.isLetter(c)) {
if (Math.random() < (toChoose/(double)letters)) {
c = Character.toUpperCase(c);
toChoose--;
}
letters--;
}
sb.append(c);
}
return sb.toString();
}
This code performs a random sample "on the fly", considering only alpha chars, as required. Use p=0.5 to switch exactly half of the letters.

Here is the code snippet for random sample problem (thanks Eyal for naming it). Not sure if that is what you are looking for.
Be aware, that this solution would run into an infinete loop if not enough lowercase letters are in the string. So you would need to tackle that as well, but I guess it is a starting point. ;-)
String myString = "9aie3ra3nr23rr5r21t";
System.out.println(upperCaseRandom(myString, 10));
public static String upperCaseRandom(String input, int n) {
StringBuilder output = new StringBuilder(input);
Random r = new Random();
for (int i = 0; i < n; i++) {
// Pick a place
int position = r.nextInt(input.length());
// Check if lowercase alpha
if (Character.isLowerCase(output.charAt(position))) {
output.setCharAt(position, Character.toUpperCase(output.charAt(position)));
} else {
i--;
}
}
return output.toString();
}
Edit:
Here is an improved version. It does change exactly n lowercase letters into uppercase letters (if there are enough, otherwise it changes all of them). The programm does not run into infinite loops, but still running time is a problem though.
public static String upperCaseRandom(String input, int n) {
final int length = input.length();
final StringBuilder output = new StringBuilder(input);
final boolean[] alreadyChecked = new boolean[length];
final Random r = new Random();
for (int i = 0, checks = 0; i < n && checks < length; i++) {
// Pick a place
int position = r.nextInt(length);
// Check if lowercase alpha
if (!alreadyChecked[position]) {
if (Character.isLowerCase(output.charAt(position))) {
output.setCharAt(position, Character.toUpperCase(output.charAt(position)));
} else {
i--;
}
checks++;
alreadyChecked[position] = true;
} else {
i--;
}
}
return output.toString();
}

I tried with
String lowerCasedRandomString = "4210281f-76ac-96b5-ed54-5458abf788d0";
String upperCasedRandomString = "4210281F-76AC-96B5-ED54-5458ABF788D0";
System.out.println(lowerCasedRandomString.toUpperCase());
System.out.println(upperCasedRandomString.toLowerCase());
I got the output
4210281F-76AC-96B5-ED54-5458ABF788D0
4210281f-76ac-96b5-ed54-5458abf788d0