Related
find the subarrays that matches the given sum input under conditions
only one loop is allowed
cant repeat the same index for sum
no inbuild methods allowed
my Output:
int arr[] = { 2, 2, 1, 4, 1, 6, 8, 5, 1, 2, 1, 1, 1 };
int checkval = 3;
results:
the subarrays are:
index 0 to 1
index 3
index 3
index 3 to 3
index 8 to 10
index 9 to 11
Subarray with most elements is 3
Kindly help me to avoid duplicate printing of single index values
also find solution to not to use the same index for next sum,
incorrect
index 8 to 10
index 9 to 11
correct
index 8 to 10
**next index should start with 11 not 9**
index 9 to 11
public class Main {
public static void main(String[] args) {
int arr[] = { 2, 2, 1, 4, 1, 6, 8, 5, 1, 2, 1, 1, 1 };
int sum = 0;
int indexval = 0;
int checkval = 4;
int i = 0;
int len[] = new int[10];
int l = 0;
int lm = 0;
System.out.println("the subarrays are: ");
while (indexval <= arr.length) {
int temp=i;
if (i < arr.length)
sum += arr[i];
else
break;
if (sum <= checkval) {
if (sum == checkval) {
System.out.println("index " + indexval + " to " + i);
len[l++] = i - indexval + 1;
lm = (i - indexval + 1) > len[l] ? (i - indexval + 1) : len[l];
i = indexval;
indexval += 1;
sum = 0;
}
}
else if (arr[temp] == checkval) {
System.out.println("index " + i);
i = indexval;
indexval += 1;
sum = 0;
}
else if (sum > checkval) {
i = indexval;
indexval += 1;
sum = 0;
}
i += 1;
}
System.out.println("Subarray with most elements is " + lm);
}
}
I have a list of 16 numbers, and I want them set as variables n1-n16. I'm looking for something similar to How to assign each element of a list to a separate variable?, but for java.
My code is:
public void setVar()
{
//x = 16;
cardNumR = cardNum;
reversed = 0;
while(cardNumR != 0) {
digit = cardNumR % 10;
reversed = reversed * 10 + digit;
cardNumR /= 10;
}
ArrayList<Long> nums = new ArrayList<Long>(x);
for (int x = 16; x > 0; x--)
{
nums.add(reversed % 10);
reversed -= (reversed % 10);
reversed /= 10;
length = nums.size();
if (length == 16)
{
System.out.println(nums);
}
}
}
which gets me a result of:
[1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6]
I want to take each of these elements and set n1 = 1, n2 = 2, n3 = 3 so on and so forth such that n16 = 6. I want to be able to do this at the end:
(2*n1) + n2 + (2*n3) + n4 + (2*n5) + n6 + (2*n7) + n8 + (2*n9) + n10 + (2*n11) + n12 + (2*n13) + n14 + (2*n15) + n16
Is there a way I can do this with a loop so that I don't have to do it one by one?
You don't need so many variables. Use a loop instead:
int sum = 0;
for(int i = 0; i < nums.size(); i++) {
if(i % 2 == 0) {
// odd index
sum += 2 * nums.get(i);
} else {
// even index
sum += nums.get(i);
}
}
Well I would just use an array
int [] array = {1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6};
or a list for that:
List<Integer> nums = List.of(1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6);
and apply the following formula:
Arrays
int result = 0;
for(int i = 0; i < array.length; i++) {
result += (i % 2 == 0) ? 2 * array[i] : array[i];
}
or if you want to avoid the use of the modulus (%) operation:
int result = 0
for(int i = 0; i < array.length; i+=2)
result += 2 * array[i];
for(int i = 1; i < array.length; i+=2)
result += array[i];
Lists
int result = 0;
for(int i = 0; i < nums.size(); i++) {
result += (i % 2 == 0) ? 2 * nums.get(i) : nums.get(i);
}
without the modulus operation :
int result = 0
for(int i = 0; i < nums.size(); i+=2)
result += 2 * nums.get(i);
for(int i = 1; i < nums.size(); i+=2)
result += nums.get(i);
Java Streams:
int result = IntStream.range(0, nums.size())
.mapToLong(i -> (i % 2 == 0) ? 2 * nums.get(i) : nums.get(i))
.sum();
I am trying to do this exercise:
Write a program that asks the user for N and M and adds up the
integers between N and M using the formula
SUM(N to M) = SUM( 1 to M ) - SUM( 1 to N-1 )
I can get this to work for positive numbers but not negative numbers.
static int method2(int n, int m) {
int sum = 0;
int sum2 = 0;
for (int i = 1; i <= m; i++) {
sum = sum + i;
}
for (int i = 1; i <= n - 1; i++) {
sum2 = sum2 + i;
}
System.out.println("sum: " + sum + ", sum2: " + sum2);
return sum = sum - sum2;
}
e.g.
using n = -1, m = 1 returns sum = 1.
Using n = -5, m = 5 returns sum = 15.
Using n = 5, m = -5 returns sum = -10.
These should all return 0.
e.g.
Using n = -2, m = 3, returns sum = 6.
Using n = -2, m = 4, returns sum = 10.
The problem is with for (int i = 1; i <= n - 1; i++), specifically i <= n - 1 because when n-1 <= 0 this will not run. I just can't think of a way around it.
Your formula
SUM(N to M) = SUM( 1 to M ) - SUM( 1 to N-1 )
Doesn't really make sense for negative values. If you give that up you can make your program simpler. We very often start for loops at 0 or 1 but that doesn't have to be the case. You could instead start your loop at a n which might be negative:
static int method2(int n, int m) {
int sum = 0;
for (int i = n; i <= m; i++) {
sum = sum + i;
}
System.out.println("sum: " + sum);
return sum;
}
You could always check before if n < 0.
And then do another reverse loop for negative numbers.
e.g.
int sum = 0;
if(m < 0){
for(int i = 0; i >= m; i--) {
sum += i;
}
} else {
for (int i = 1; i <= m; i++) {
sum += i;
}
}
If you really have to use that formula you could use instead of:
for (int i = 1; i <= m; i++) {
the following code which changes the index either by 1 or by -1
for (int i = 1; i <= m; i+=(int)Math.signum(m-1+0.1)) {
(added 0.1 such that in case m is 1 the result is positive and not 0)
Ofc you should do the same for n.
I am trying to add two int arrays together to get a sum. The first array contains 0000000000000000123456789 and the second array contains 0001111111111111111111111. The sum should be 1111111111111234567900. Also I'm trying to do this without BigInteger or BigDecimal.
for(int i=0; i<number1.length;i++){
add= number1[i]+number2[i];
if(plus>=10){
sum[i-1]+=add/10;
sum[i] = add%10;
}else{
sum[i]=add;
}
}
The output that is produced at the moment is 00011111111111112345678100. How can I fix my code so that the 8 becomes a 9?
This kinda works. I can think of a couple of cases where something like this would break, like if the arrays were like {9,9,9} and {9,9,9}, result would be {9,9,8} instead of {1,9,9,8}. It's a minor fix that is being left as an activity to the reader.
public static void main(String []args){
int[] number1 = {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,2,3,4,5,6,7,8,9};
int[] number2 = {0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1};
int carry = 0, sum = 0;
int[] result = new int[number1.length];
for (int i = number1.length - 1; i >= 0 ; i--) {
sum = number1[i] + number2[i] + carry;
result[i] = sum%10;
carry = sum/10;
}
// [0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 3, 4, 5, 6, 7, 9, 0, 0]
System.out.println(Arrays.toString(result));
}
Try this
int[] number = {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,2,3,4,5,6,7,8,9};
int[] number2 = {0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1};
int[] result = new int[number.length+1];
int carry = 0;
for (int i = number.length - 1; i >= 0; i--) {
result[i+1] = number[i] + number2[i] + carry;
if (result[i+1] > 9) {
carry = result[i+1] - 9;
result[i+1] -= 10;
}
else {
carry = 0;
}
}
System.out.println(Arrays.toString(result));
This code should work for properly carrying numbers greater than 10 to the next up and keep the lower numbers!
int[] result = new int[number1.length+1];
for (int i=number1.length-1; i>=0;i--){
result[i+1] = number1[i] + number[i] + result[i+1];
if(result[i+1] >= 10){
result[i+1] -= 10; // it could never be more than 20 so this is ok and if it is 10 than it will still carry over to the next
result[i] = 1; // the ten carried to the next one
}
}
Similar to Seek Addo's answer, but avoids the errors found in his code that prevent the application from properly carrying the numbers.
This is the most compact way I could think to do it, and should yield the proper answers of
[1,1,1,1,1,1,1,1,1,1,1,1,1,2,3,4,5,6,7,9,0,0] or 1111111111111234567900
The accepted answer is not good. This is my version of it (support for different array sizes):
public static int[] sumIntArrays(int[] n1, int[] n2) {
int n1Size = n1.length - 1, n2Size = n2.length - 1, indexSum = n1Size > n2Size ? n1Size + 2 : n2Size + 2, slack = 0, s;
int[] sum = new int[indexSum];
while (true) {
if (n1Size < 0 && n2Size < 0)
break;
s = (n1Size < 0 ? 0 : n1[n1Size--]) + (n2Size < 0 ? 0 : n2[n2Size--]) + slack;
if (s > 9) {
sum[--indexSum] = s % 10;
slack = s / 10;
} else {
sum[--indexSum] = s;
slack = 0;
}
}
if (slack != 0)
sum[0] = slack;
return sum;
}
We know that all primes above 3 can be generated using:
6 * k + 1
6 * k - 1
However we all numbers generated from the above formulas are not prime.
For Example:
6 * 6 - 1 = 35 which is clearly divisible by 5.
To Eliminate such conditions, I used a Sieve Method and removing the numbers which are factors of the numbers generated from the above formula.
Using the facts:
A number is said to be prime if it has no prime factors.
As we can generate all the prime numbers using the above formulas.
If we can remove all the multiples of the above numbers we are left with only prime numbers.
To generate primes below 1000.
ArrayList<Integer> primes = new ArrayList<>();
primes.add(2);//explicitly add
primes.add(3);//2 and 3
int n = 1000;
for (int i = 1; i <= (n / 6) ; i++) {
//get all the numbers which can be generated by the formula
int prod6k = 6 * i;
primes.add(prod6k - 1);
primes.add(prod6k + 1);
}
for (int i = 0; i < primes.size(); i++) {
int k = primes.get(i);
//remove all the factors of the numbers generated by the formula
for(int j = k * k; j <= n; j += k)//changed to k * k from 2 * k, Thanks to DTing
{
int index = primes.indexOf(j);
if(index != -1)
primes.remove(index);
}
}
System.out.println(primes);
However, this method does generate the prime numbers correctly. This runs in a much faster way as we need not check for all the numbers which we do check in a Sieve.
My question is that am I missing any edge case? This would be a lot better but I never saw someone using this. Am I doing something wrong?
Can this approach be much more optimized?
Taking a boolean[] instead of an ArrayList is much faster.
int n = 100000000;
boolean[] primes = new boolean[n + 1];
for (int i = 0; i <= n; i++)
primes[i] = false;
primes[2] = primes[3] = true;
for (int i = 1; i <= n / 6; i++) {
int prod6k = 6 * i;
primes[prod6k + 1] = true;
primes[prod6k - 1] = true;
}
for (int i = 0; i <= n; i++) {
if (primes[i]) {
int k = i;
for (int j = k * k; j <= n && j > 0; j += k) {
primes[j] = false;
}
}
}
for (int i = 0; i <= n; i++)
if (primes[i])
System.out.print(i + " ");
5 is the first number generated by your criteria. Let's take a look at the numbers generated up to 25:
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25
Now, let's look at these same numbers, when we use the Sieve of Eratosthenes algorithm:
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25
After removing 2:
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25
After removing 3:
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25
This is the same as the first set! Notice they both include 25, which is not prime. If we think about it, this is an obvious result. Consider any group of 6 consecutive numbers:
6k - 3, 6k - 2, 6k - 1, 6k, 6k + 1, 6k + 2
If we factor a little, we get:
3*(2k - 1), 2*(3k - 1), 6k - 1, 6*(k), 6k + 1, 2*(3k + 1)
In any group of 6 consecutive numbers, three of them will be divisible by two, and two of them will be divisible by three. These are exactly the numbers we have removed so far! Therefore:
Your algorithm to only use 6k - 1 and 6k + 1 is exactly the same as the first two rounds of the Sieve of Erathosthenes.
It's a pretty nice speed improvement over the Sieve, too, because we don't have to add all those extra elements just to remove them. This explains why your algorithm works and why it doesn't miss any cases; because it's exactly the same as the Sieve.
Anyway, I agree that once you've generated primes, your boolean way is by far the fastest. I have set up a benchmark using your ArrayList way, your boolean[] way, and my own way using LinkedList and iterator.remove() (because removals are fast in a LinkedList. Here's the code for my test harness. Note that I run the test 12 times to ensure that the JVM is warmed up, and I print the size of the list and change the size of n to attempt to prevent too much branch prediction optimization. You can also get faster in all three methods by using += 6 in the initial seed, instead of prod6k:
import java.util.*;
public class PrimeGenerator {
public static List<Integer> generatePrimesArrayList(int n) {
List<Integer> primes = new ArrayList<>(getApproximateSize(n));
primes.add(2);// explicitly add
primes.add(3);// 2 and 3
for (int i = 6; i <= n; i+=6) {
// get all the numbers which can be generated by the formula
primes.add(i - 1);
primes.add(i + 1);
}
for (int i = 0; i < primes.size(); i++) {
int k = primes.get(i);
// remove all the factors of the numbers generated by the formula
for (int j = k * k; j <= n; j += k)// changed to k * k from 2 * k, Thanks
// to DTing
{
int index = primes.indexOf(j);
if (index != -1)
primes.remove(index);
}
}
return primes;
}
public static List<Integer> generatePrimesBoolean(int n) {
boolean[] primes = new boolean[n + 5];
for (int i = 0; i <= n; i++)
primes[i] = false;
primes[2] = primes[3] = true;
for (int i = 6; i <= n; i+=6) {
primes[i + 1] = true;
primes[i - 1] = true;
}
for (int i = 0; i <= n; i++) {
if (primes[i]) {
int k = i;
for (int j = k * k; j <= n && j > 0; j += k) {
primes[j] = false;
}
}
}
int approximateSize = getApproximateSize(n);
List<Integer> primesList = new ArrayList<>(approximateSize);
for (int i = 0; i <= n; i++)
if (primes[i])
primesList.add(i);
return primesList;
}
private static int getApproximateSize(int n) {
// Prime Number Theorem. Round up
int approximateSize = (int) Math.ceil(((double) n) / (Math.log(n)));
return approximateSize;
}
public static List<Integer> generatePrimesLinkedList(int n) {
List<Integer> primes = new LinkedList<>();
primes.add(2);// explicitly add
primes.add(3);// 2 and 3
for (int i = 6; i <= n; i+=6) {
// get all the numbers which can be generated by the formula
primes.add(i - 1);
primes.add(i + 1);
}
for (int i = 0; i < primes.size(); i++) {
int k = primes.get(i);
for (Iterator<Integer> iterator = primes.iterator(); iterator.hasNext();) {
int primeCandidate = iterator.next();
if (primeCandidate == k)
continue; // Always skip yourself
if (primeCandidate == (primeCandidate / k) * k)
iterator.remove();
}
}
return primes;
}
public static void main(String... args) {
int initial = 4000;
for (int i = 0; i < 12; i++) {
int n = initial * i;
long start = System.currentTimeMillis();
List<Integer> result = generatePrimesArrayList(n);
long seconds = System.currentTimeMillis() - start;
System.out.println(result.size() + "\tArrayList Seconds: " + seconds);
start = System.currentTimeMillis();
result = generatePrimesBoolean(n);
seconds = System.currentTimeMillis() - start;
System.out.println(result.size() + "\tBoolean Seconds: " + seconds);
start = System.currentTimeMillis();
result = generatePrimesLinkedList(n);
seconds = System.currentTimeMillis() - start;
System.out.println(result.size() + "\tLinkedList Seconds: " + seconds);
}
}
}
And the results of the last few trials:
3432 ArrayList Seconds: 430
3432 Boolean Seconds: 0
3432 LinkedList Seconds: 90
3825 ArrayList Seconds: 538
3824 Boolean Seconds: 0
3824 LinkedList Seconds: 81
4203 ArrayList Seconds: 681
4203 Boolean Seconds: 0
4203 LinkedList Seconds: 100
4579 ArrayList Seconds: 840
4579 Boolean Seconds: 0
4579 LinkedList Seconds: 111
You don't need to add all possible candidates to the array. You can create a Set to store all non primes.
Also you can start checking at k * k, rather than 2 * k
public void primesTo1000() {
Set<Integer> notPrimes = new HashSet<>();
ArrayList<Integer> primes = new ArrayList<>();
primes.add(2);//explicitly add
primes.add(3);//2 and 3
for (int i = 1; i < (1000 / 6); i++) {
handlePossiblePrime(6 * i - 1, primes, notPrimes);
handlePossiblePrime(6 * i + 1, primes, notPrimes);
}
System.out.println(primes);
}
public void handlePossiblePrime(
int k, List<Integer> primes, Set<Integer> notPrimes) {
if (!notPrimes.contains(k)) {
primes.add(k);
for (int j = k * k; j <= 1000; j += k) {
notPrimes.add(j);
}
}
}
untested code, check corners
Here is a bit packing version of the sieve as suggested in the answer referenced by #Will Ness. Rather than return the nth prime, this version returns a list of primes to n:
public List<Integer> primesTo(int n) {
List<Integer> primes = new ArrayList<>();
if (n > 1) {
int limit = (n - 3) >> 1;
int[] sieve = new int[(limit >> 5) + 1];
for (int i = 0; i <= (int) (Math.sqrt(n) - 3) >> 1; i++)
if ((sieve[i >> 5] & (1 << (i & 31))) == 0) {
int p = i + i + 3;
for (int j = (p * p - 3) >> 1; j <= limit; j += p)
sieve[j >> 5] |= 1 << (j & 31);
}
primes.add(2);
for (int i = 0; i <= limit; i++)
if ((sieve[i >> 5] & (1 << (i & 31))) == 0)
primes.add(i + i + 3);
}
return primes;
}
There seems to be a bug in your updated code that uses a boolean array (it is not returning all the primes).
public static List<Integer> booleanSieve(int n) {
boolean[] primes = new boolean[n + 5];
for (int i = 0; i <= n; i++)
primes[i] = false;
primes[2] = primes[3] = true;
for (int i = 1; i <= n / 6; i++) {
int prod6k = 6 * i;
primes[prod6k + 1] = true;
primes[prod6k - 1] = true;
}
for (int i = 0; i <= n; i++) {
if (primes[i]) {
int k = i;
for (int j = k * k; j <= n && j > 0; j += k) {
primes[j] = false;
}
}
}
List<Integer> primesList = new ArrayList<>();
for (int i = 0; i <= n; i++)
if (primes[i])
primesList.add(i);
return primesList;
}
public static List<Integer> bitPacking(int n) {
List<Integer> primes = new ArrayList<>();
if (n > 1) {
int limit = (n - 3) >> 1;
int[] sieve = new int[(limit >> 5) + 1];
for (int i = 0; i <= (int) (Math.sqrt(n) - 3) >> 1; i++)
if ((sieve[i >> 5] & (1 << (i & 31))) == 0) {
int p = i + i + 3;
for (int j = (p * p - 3) >> 1; j <= limit; j += p)
sieve[j >> 5] |= 1 << (j & 31);
}
primes.add(2);
for (int i = 0; i <= limit; i++)
if ((sieve[i >> 5] & (1 << (i & 31))) == 0)
primes.add(i + i + 3);
}
return primes;
}
public static void main(String... args) {
Executor executor = Executors.newSingleThreadExecutor();
executor.execute(() -> {
for (int i = 0; i < 10; i++) {
int n = (int) Math.pow(10, i);
Stopwatch timer = Stopwatch.createUnstarted();
timer.start();
List<Integer> result = booleanSieve(n);
timer.stop();
System.out.println(result.size() + "\tBoolean: " + timer);
}
for (int i = 0; i < 10; i++) {
int n = (int) Math.pow(10, i);
Stopwatch timer = Stopwatch.createUnstarted();
timer.start();
List<Integer> result = bitPacking(n);
timer.stop();
System.out.println(result.size() + "\tBitPacking: " + timer);
}
});
}
0 Boolean: 38.51 μs
4 Boolean: 45.77 μs
25 Boolean: 31.56 μs
168 Boolean: 227.1 μs
1229 Boolean: 1.395 ms
9592 Boolean: 4.289 ms
78491 Boolean: 25.96 ms
664116 Boolean: 133.5 ms
5717622 Boolean: 3.216 s
46707218 Boolean: 32.18 s
0 BitPacking: 117.0 μs
4 BitPacking: 11.25 μs
25 BitPacking: 11.53 μs
168 BitPacking: 70.03 μs
1229 BitPacking: 471.8 μs
9592 BitPacking: 3.701 ms
78498 BitPacking: 9.651 ms
664579 BitPacking: 43.43 ms
5761455 BitPacking: 1.483 s
50847534 BitPacking: 17.71 s
There are several things that could be optimized.
For starters, the "contains" and "removeAll" operations on an ArrayList are rather expensive operations (linear for the former, worst case quadratic for the latter) so you might not want to use the ArrayList for this. A Hash- or TreeSet has better complexities for this, being nearly constant (Hashing complexities are weird) and logarithmic I think
You could look into the sieve of sieve of Eratosthenes if you want a more efficient sieve altogeter, but that would be besides the point of your question about the 6k +-1 trick. It is slightly but not noticably more memory expensive than your solution, but way faster.
Can this approach be much more optimized?
The answer is yes.
I'll start by saying that it is a good idea to use the sieve on a subset of number within a certain range, and your suggesting is doing exactly that.
Reading about generating Primes:
...Furthermore, based on the sieve formalisms, some integer sequences
(sequence A240673 in OEIS) are constructed which they also could be used for generating primes in certain intervals.
The meaning of this paragraph is that your approach of starting with a reduced list of integers was indeed adopted by the academy, but their techniques are more efficient (but also, naturally, more complex).
You can generate your trial numbers with a wheel, adding 2 and 4 alternately, that eliminates the multiplication in 6 * k +/- 1.
public void primesTo1000() {
Set<Integer> notPrimes = new HashSet<>();
ArrayList<Integer> primes = new ArrayList<>();
primes.add(2); //explicitly add
primes.add(3); //2 and 3
int step = 2;
int num = 5 // 2 and 3 already handled.
while (num < 1000) {
handlePossiblePrime(num, primes, notPrimes);
num += step; // Step to next number.
step = 6 - step; // Step by 2, 4 alternately.
}
System.out.println(primes);
}
Probably the most suitable standard datastructure for Sieve of Eratosthenes is the BitSet. Here's my solution:
static BitSet genPrimes(int n) {
BitSet primes = new BitSet(n);
primes.set(2); // add 2 explicitly
primes.set(3); // add 3 explicitly
for (int i = 6; i <= n ; i += 6) { // step by 6 instead of multiplication
primes.set(i - 1);
primes.set(i + 1);
}
int max = (int) Math.sqrt(n); // don't need to filter multiples of primes bigger than max
// this for loop enumerates all set bits starting from 5 till the max
// sieving 2 and 3 is meaningless: n*6+1 and n*6-1 are never divisible by 2 or 3
for (int i = primes.nextSetBit(5); i >= 0 && i <= max; i = primes.nextSetBit(i+1)) {
// The actual sieve algorithm like in your code
for(int j = i * i; j <= n; j += i)
primes.clear(j);
}
return primes;
}
Usage:
BitSet primes = genPrimes(1000); // generate primes up to 1000
System.out.println(primes.cardinality()); // print number of primes
// print all primes like {2, 3, 5, ...}
System.out.println(primes);
// print all primes one per line
for(int prime = primes.nextSetBit(0); prime >= 0; prime = primes.nextSetBit(prime+1))
System.out.println(prime);
// print all primes one per line using java 8:
primes.stream().forEach(System.out::println);
The boolean-based version may work faster for small n values, but if you need, for example, a million of prime numbers, BitSet will outperform it in several times and actually works correctly. Here's lame benchmark:
public static void main(String... args) {
long start = System.nanoTime();
BitSet res = genPrimes(10000000);
long diff = System.nanoTime() - start;
System.out.println(res.cardinality() + "\tBitSet Seconds: " + diff / 1e9);
start = System.nanoTime();
List<Integer> result = generatePrimesBoolean(10000000); // from durron597 answer
diff = System.nanoTime() - start;
System.out.println(result.size() + "\tBoolean Seconds: " + diff / 1e9);
}
Output:
664579 BitSet Seconds: 0.065987717
664116 Boolean Seconds: 0.167620323
664579 is the correct number of primes below 10000000.
This method below shows how to find prime nos using 6k+/-1 logic
this was written in python 3.6
def isPrime(n):
if(n<=1):
return 0
elif(n<4): #2 , 3 are prime
return 1
elif(n%2==0): #already excluded no.2 ,so any no. div. by 2 cant be prime
return 0
elif(n<9): #5, 7 are prime and 6,8 are excl. in the above step
return 1
elif(n%3==0):
return 1
f=5 #Till now we have checked the div. of n with 2,3 which means with 4,6,8 also now that is why f=5
r=int(n**.5) #rounding of root n, i.e: floor(sqrt(n)) r*r<=n
while(f<=r):
if(n%f==0): #checking if n has any primefactor lessthan sqrt(n), refer LINE 1
return 0
if(n%(f+2)==0): #remember her we are not incrementing f, see the 6k+1 rule to understand this while loop steps ,you will see that most values of f are prime
return 0
f=f+6
return 1
def prime_nos():
counter=2 #we know 2,3 are prime
print(2)
print(3) #we know 2,3 are prime
i=1
s=5 #sum 2+3
t=0
n=int(input("Enter the upper limit( should be > 3: "))
n=(n-1)//6 #finding max. limit(n=6*i+1) upto which I(here n on left hand side) should run
while(i<n):#2*(10**6)):
if (isPrime(6*i-1)):
counter=counter+1
print(6*i-1) #prime no
if(isPrime(6*i+1)):
counter=counter+1
print(6*i+1) #prime no
i+=1
prime_nos() #fn. call
Your prime number formula mathematically incorrect ex. take 96 it dividable to 6 96/6=16 so by this logic 97 and 95 must be prime if square root passed but square root of 95 is 9.7467... (passed) so its "prime". But 95 clearly dividable by 5 fast algorithm in c#
int n=100000000;
bool [] falseprimes = new bool[n + 2];
int ed=n/6;
ed = ed * 6;
int md = (int)Math.Sqrt((double)ed);
for (int i = ed; i > md; i-=6)
{
falseprimes[i + 1] = true;
falseprimes[i - 1] = true;
}
md = md / 6;
md = md * 5;
for (int i = md; i > 5; i -= 6)
{
falseprimes[i + 1] = true;
falseprimes[i - 1] = true;
falseprimes[(i + 1)* (i + 1)] = false;
falseprimes[(i-1) * (i-1)] = false;
}
falseprimes[2] = true;
falseprimes[3] = true;
To generate prime numbers using 6 * k + - 1 rule use this algorithm:
int n = 100000000;
int j,jmax=n/6;
boolean[] primes5m6 = new boolean[jmax+1];
boolean[] primes1m6 = new boolean[jmax+1];
for (int i = 0; i <= jmax; i++){
primes5m6[i] = false;
primes1m6[i] = false;
}
for (int i = 1; i <= (int)((Math.sqrt(n)+1)/6)+1; i++){
if (!primes5m6[i]){
for (j = 6*i*i; j <= jmax; j+=6*i-1){
primes5m6[j]=true;
primes1m6[j-2*i]=true;
}
for (; j <= jmax+2*i; j+=6*i-1)
primes1m6[j-2*i]=true;
}
if (!primes1m6[i]){
for (j = 6*i*i; j <= jmax-2*i; j+=6*i+1){
primes5m6[j]=true;
primes1m6[j+2*i]=true;
}
for (; j <= jmax; j+=6*i+1)
primes5m6[j]=true;
}
}
System.out.print(2 + " ");
System.out.print(3 + " ");
for (int i = 1; i <= jmax; i++){
if (!primes5m6[i])
System.out.print((6*i-1) + " ");
if (!primes1m6[i])
System.out.print((6*i+1) + " ");
}