The sequence of triangle numbers is
generated by adding the natural
numbers. So the 7th triangle number
would be 1 + 2 + 3 + 4 + 5 + 6 + 7 =
28. The first ten terms would be:
1, 3, 6, 10, 15, 21, 28, 36, 45, 55,
...
Let us list the factors of the first
seven triangle numbers:
1: 1
3: 1,3
6: 1,2,3,6
10: 1,2,5,10
15: 1,3,5,15
21: 1,3,7,21
28: 1,2,4,7,14,28
We can see that 28 is the first
triangle number to have over five
divisors.
Given an integer n, display the first
triangle number having at least n
divisors.
Sample Input: 5
Output 28
Input Constraints: 1<=n<=320
I was obviously able to do this question, but I used a naive algorithm:
Get n.
Find triangle numbers and check their number of factors using the mod operator.
But the challenge was to show the output within 4 seconds of input. On high inputs like 190 and above it took almost 15-16 seconds. Then I tried to put the triangle numbers and their number of factors in a 2d array first and then get the input from the user and search the array. But somehow I couldn't do it: I got a lot of processor faults. Please try doing it with this method and paste the code. Or if there are any better ways, please tell me.
Here's a hint:
The number of divisors according to the Divisor function is the product of the power of each prime factor plus 1. For example, let's consider the exponential prime representation of 28:
28 = 22 * 30 * 50 * 71 * 110...
The product of each exponent plus one is: (2+1)*(0+1)*(0+1)*(1+1)*(0+1)... = 6, and sure enough, 28 has 6 divisors.
Now, consider that the nth triangular number can be computed in closed form as n(n+1)/2. We can multiply numbers written in the exponential prime form simply by adding up the exponents at each position. Dividing by two just means decrementing the exponent on the two's place.
Do you see where I'm going with this?
Well, you don't go into a lot of detail about what you did, but I can give you an optimization that can be used, if you didn't think of it...
If you're using the straightforward method of trying to find factors of a number n, by using the mod operator, you don't need to check all the numbers < n. That obviously would take n comparisons...you can just go up to floor(sqrt(n)). For each factor you find, just divide n by that number, and you'll get the conjugate value, and not need to find it manually.
For example: say n is 15.
We loop, and try 1 first. Yep, the mod checks out, so it's a factor. We divide n by the factor to get the conjugate value, so we do (15 / 1) = 15...so 15 is a factor.
We try 2 next. Nope. Then 3. Yep, which also gives us (15 / 3) = 5.
And we're done, because 4 is > floor(sqrt(n)). Quick!
If you didn't think of it, that might be something you could leverage to improve your times...overall you go from O(n) to O(sqrt (n)) which is pretty good (though for numbers this small, constants may still weigh heavily.)
I was in a programming competition way back in school where there was some similar question with a run time limit. the team that "solved" it did as follows:
1) solve it with a brute force slow method.
2) write a program to just print out the answer (you found using the slow method), which will run sub second.
I thought this was bogus, but they won.
see Triangular numbers: a(n) = C(n+1,2) = n(n+1)/2 = 0+1+2+...+n. (Formerly M2535 N1002)
then pick the language you want implement it in, see this:
"... Python
import math
def diminishing_returns(val, scale):
if val < 0:
return -diminishing_returns(-val, scale)
mult = val / float(scale)
trinum = (math.sqrt(8.0 * mult + 1.0) - 1.0) / 2.0
return trinum * scale
..."
First, create table with two columns: Triangle_Number Count_of_Factors.
Second, derive from this a table with the same columns, but consisting only of the 320 rows of the lowest triangle number with a distinct number of factors.
Perform your speedy lookup to the second table.
If you solved the problem, you should be able to access the thread on Project Euler in which people post their (some very efficient) solutions.
If you're going to copy and paste a problem, please cite the source (unless it was your teacher who stole it); and I second Wouter van Niferick's comment.
Well, at least you got a good professor. Performance is important.
Since you have a program that can do the job, you can precalculate all of the answers for 1 .. 320.
Store them in an array, then simply subscript into the array to get the answer. That will be very fast.
Compile with care, winner of worst code of the year :D
#include <iostream>
bool isPrime( unsigned long long number ){
if( number != 2 && number % 2 == 0 )
return false;
for( int i = 3;
i < static_cast<unsigned long long>
( sqrt(static_cast<double>(number)) + 1 )
; i += 2 ){
if( number % i == 0 )
return false;
}
return true;
}
unsigned int p;
unsigned long long primes[1024];
void initPrimes(){
primes[0] = 2;
primes[1] = 3;
unsigned long long number = 5;
for( unsigned int i = 2; i < 1024; i++ ){
while( !isPrime(number) )
number += 2;
primes[i] = number;
number += 2;
}
return;
}
unsigned long long nextPrime(){
unsigned int ret = p;
p++;
return primes[ret];
}
unsigned long long numOfDivs( unsigned long long number ){
p = 0;
std::vector<unsigned long long> v;
unsigned long long prime = nextPrime(), divs = 1, i = 0;
while( number >= prime ){
i = 0;
while( number % prime == 0 ){
number /= prime;
i++;
}
if( i )
v.push_back( i );
prime = nextPrime();
}
for( unsigned n = 0; n < v.size(); n++ )
divs *= (v[n] + 1);
return divs;
}
unsigned long long nextTriNumber(){
static unsigned long long triNumber = 1, next = 2;
unsigned long long retTri = triNumber;
triNumber += next;
next++;
return retTri;
}
int main()
{
initPrimes();
unsigned long long n = nextTriNumber();
unsigned long long divs = 500;
while( numOfDivs(n) <= divs )
n = nextTriNumber();
std::cout << n;
std::cin.get();
}
def first_triangle_number_with_over_N_divisors(N):
n = 4
primes = [2, 3]
fact = [None, None, {2:1}, {3:1}]
def num_divisors (x):
num = 1
for mul in fact[x].values():
num *= (mul+1)
return num
while True:
factn = {}
for p in primes:
if p > n//2: break
r = n // p
if r * p == n:
factn = fact[r].copy()
factn[p] = factn.get(p,0) + 1
if len(factn)==0:
primes.append(n)
factn[n] = 1
fact.append(factn)
(x, y) = (n-1, n//2) if n % 2 == 0 else (n, (n-1)//2)
numdiv = num_divisors(x) * num_divisors(y)
if numdiv >= N:
print('Triangle number %d: %d divisors'
%(x*y, numdiv))
break
n += 1
>>> first_triangle_number_with_over_N_divisors(500)
Triangle number 76576500: 576 divisors
Dude here is ur code, go have a look. It calculates the first number that has divisors greater than 500.
void main() {
long long divisors = 0;
long long nat_num = 0;
long long tri_num = 0;
int tri_sqrt = 0;
while (1) {
divisors = 0;
nat_num++;
tri_num = nat_num + tri_num;
tri_sqrt = floor(sqrt((double)tri_num));
long long i = 0;
for ( i=tri_sqrt; i>=1; i--) {
long long remainder = tri_num % i;
if ( remainder == 0 && tri_num == 1 ) {
divisors++;
}
else if (remainder == 0 && tri_num != 1) {
divisors++;
divisors++;
}
}
if (divisors >100) {
cout <<"No. of divisors: "<<divisors<<endl<<tri_num<<endl;
}
if (divisors > 500)
break;
}
cout<<"Final Result: "<<tri_num<<endl;
system("pause");
}
Boojum's answer motivated me to write this little program. It seems to work well, although it does use a brute force method of computing primes. It's neat how all the natural numbers can be broken down into prime number components.
#include <stdio.h>
#include <stdlib.h>
#include <iostream>
#include <iomanip>
#include <vector>
//////////////////////////////////////////////////////////////////////////////
typedef std::vector<size_t> uint_vector;
//////////////////////////////////////////////////////////////////////////////
// add a prime number to primes[]
void
primeAdd(uint_vector& primes)
{
size_t n;
if (primes.empty())
{
primes.push_back(2);
return;
}
for (n = *(--primes.end()) + 1; ; ++n)
{
// n is even -> not prime
if ((n & 1) == 0) continue;
// look for a divisor in [2,n)
for (size_t i = 2; i < n; ++i)
{
if ((n % i) == 0) continue;
}
// found a prime
break;
}
primes.push_back(n);
}
//////////////////////////////////////////////////////////////////////////////
void
primeFactorize(size_t n, uint_vector& primes, uint_vector& f)
{
f.clear();
for (size_t i = 0; n > 1; ++i)
{
while (primes.size() <= i) primeAdd(primes);
while (f.size() <= i) f.push_back(0);
while ((n % primes[i]) == 0)
{
++f[i];
n /= primes[i];
}
}
}
//////////////////////////////////////////////////////////////////////////////
int
main(int argc, char** argv)
{
// allow specifying number of TN's to be evaluated
size_t lim = 1000;
if (argc > 1)
{
lim = atoi(argv[1]);
}
if (lim == 0) lim = 1000;
// prime numbers
uint_vector primes;
// factors of (n), (n + 1)
uint_vector* f = new uint_vector();
uint_vector* f1 = new uint_vector();
// sum vector
uint_vector sum;
// prime factorize (n)
size_t n = 1;
primeFactorize(n, primes, *f);
// iterate over triangle-numbers
for (; n <= lim; ++n)
{
// prime factorize (n + 1)
primeFactorize(n + 1, primes, *f1);
while (f->size() < f1->size()) f->push_back(0);
while (f1->size() < f->size()) f1->push_back(0);
size_t numTerms = f->size();
// compute prime factors for (n * (n + 1) / 2)
sum.clear();
size_t i;
for (i = 0; i < numTerms; ++i)
{
sum.push_back((*f)[i] + (*f1)[i]);
}
--sum[0];
size_t numFactors = 1, tn = 1;
for (i = 0; i < numTerms; ++i)
{
size_t exp = sum[i];
numFactors *= (exp + 1);
while (exp-- != 0) tn *= primes[i];
}
std::cout
<< n << ". Triangle number "
<< tn << " has " << numFactors << " factors."
<< std::endl;
// prepare for next iteration
f->clear();
uint_vector* tmp = f;
f = f1;
f1 = tmp;
}
delete f;
delete f1;
return 0;
}
Related
I should implement this summation in C ++. I have tried with this code, but with very high numbers up to 10 ^ 12 it takes too long.
The summation is:
For any positive integer k, let d(k) denote the number of positive divisors of k (including 1 and k itself).
For example, for the number 4: 1 has 1 divisor, 2 has two divisors, 3 has two divisors, and 4 has three divisors. So the result would be 8.
This is my code:
#include <iostream>
#include <algorithm>
using namespace std;
int findDivisors(long long n)
{
int c=0;
for(int j=1;j*j<=n;j++)
{
if(n%j==0)
{
c++;
if(j!=(n/j))
{
c++;
}
}
}
return c;
}
long long compute(long long n)
{
long long sum=0;
for(int i=1; i<=n; i++)
{
sum += (findDivisors(i));
}
return sum;
}
int main()
{
int n, divisors;
freopen("input.txt", "r", stdin);
freopen("output.txt", "w", stdout);
cin >> n;
cout << compute(n);
}
I think it's not just a simple optimization problem, but maybe I should change the algorithm entirely.
Would anyone have any ideas to speed it up? Thank you.
largest_prime_is_463035818's answer shows an O(N) solution, but the OP is trying to solve this problem
with very high numbers up to 1012.
The following is an O(N1/2) algorithm, based on some observations about the sum
n/1 + n/2 + n/3 + ... + n/n
In particular, we can count the number of terms with a specific value.
Consider all the terms n/k where k > n/2. There are n/2 of those and all are equal to 1 (integer division), so that their sum is n/2.
Similar considerations hold for the other dividends, so that we can write the following function
long long count_divisors(long long n)
{
auto sum{ n };
for (auto i{ 1ll }, k_old{ n }, k{ n }; i < k ; ++i, k_old = k)
{ // ^^^^^ it goes up to sqrt(n)
k = n / (i + 1);
sum += (k_old - k) * i;
if (i == k)
break;
sum += k;
}
return sum;
}
Here it is tested against the O(N) algorithm, the only difference in the results beeing the corner cases n = 0 and n = 1.
Edit
Thanks again to largest_prime_is_463035818, who linked the Wikipedia page about the divisor summatory function, where both an O(N) and an O(sqrt(N)) algorithm are mentioned.
An implementation of the latter may look like this
auto divisor_summatory(long long n)
{
auto sum{ 0ll };
auto k{ 1ll };
for ( ; k <= n / k; ++k )
{
sum += n / k;
}
--k;
return 2 * sum - k * k;
}
They also add this statement:
Finding a closed form for this summed expression seems to be beyond the techniques available, but it is possible to give approximations. The leading behavior of the series is given by
D(x) = xlogx + x(2γ - 1) + Δ(x)
where γ is the Euler–Mascheroni constant, and the error term is Δ(x) = O(sqrt(x)).
I used your brute force approach as reference to have test cases. The ones I used are
compute(12) == 35
cpmpute(100) == 482
Don't get confused by computing factorizations. There are some tricks one can play when factorizing numbers, but you actually don't need any of that. The solution is a plain simple O(N) loop:
#include <iostream>
#include <limits>
long long compute(long long n){
long long sum = n+1;
for (long long i=2; i < n ; ++i){
sum += n/i;
}
return sum;
}
int main()
{
std::cout << compute(12) << "\n";
std::cout << compute(100) << "\n";
}
Output:
35
482
Why does this work?
The key is in Marc Glisse's comment:
As often with this kind of problem, this sum actually counts pairs x,
y where x divides y, and the sum is arranged to count first all x
corresponding to a fixed y, but nothing says you have to keep it that
way.
I could stop here, because the comment already explains it all. Though, if it didn't click yet...
The trick is to realize that it is much simpler to count divisors of all numbers up to n rather than n-times counting divisors of individual numbers and take the sum.
You don't need to care about factorizations of eg 123123123 or 52323423 to count all divisors up to 10000000000. All you need is a change of perspective. Instead of trying to factorize numbers, consider the divisors. How often does the divisor 1 appear up to n? Simple: n-times. How often does the divisor 2 appear? Still simple: n/2 times, because every second number is divisible by 2. Divisor 3? Every 3rd number is divisible by 3. I hope you can see the pattern already.
You could even reduce the loop to only loop till n/2, because bigger numbers obviously appear only once as divisor. Though I didn't bother to go further, because the biggest change is from your O(N * sqrt(N)) to O(N).
Let's start off with some math and reduce the O(n * sq(n)) factorization to O(n * log(log(n))) and for counting the sum of divisors the overall complexity is O(n * log(log(n)) + n * n^(1/3)).
For instance:
In Codeforces himanshujaju explains how we can optimize the solution of finding divisors of a number.
I am simplifying it a little bit.
Let, n as the product of three numbers p, q, and r.
so assume p * q * r = n, where p <= q <= r.
The maximum value of p = n^(1/3).
Now we can loop over all prime numbers in a range [2, n^(1/3)]
and try to reduce the time complexity of prime factorization.
We will split our number n into two numbers x and y => x * y = n.
And x contains prime factors up to n^(1/3) and y deals with higher prime factors greater than n^(1/3).
Thus gcd(x, y) = 1.
Now define F(n) as the number of prime factors of n.
From multiplicative rules, we can say that
F(x * y) = F(x) * F(y), if gcd(x, y) = 1.
For finding F(n) => F(x * y) = F(x) * F(y)
So first find F(x) then F(y) will F(n/x)
And there will 3 cases to cover for y:
1. y is a prime number: F(y) = 2.
2. y is the square of a prime number: F(y) = 3.
3. y is a product of two distinct prime numbers: F(y) = 4.
So once we are done with finding F(x) and F(y), we are also done with finding F(x * y) or F(n).
In Cp-Algorithm there is also a nice explanation of how to count the number of divisors on a number. And also in GeeksForGeeks a nice coding example of how to count the number of divisors of a number in an efficient way. One can check the articles and can generate a nice solution to this problem.
C++ implementation
#include <bits/stdc++.h>
using namespace std;
const int maxn = 1e6 + 11;
bool prime[maxn];
bool primesquare[maxn];
int table[maxn]; // for storing primes
void SieveOfEratosthenes()
{
for(int i = 2; i < maxn; i++){
prime[i] = true;
}
for(int i = 0; i < maxn; i++){
primesquare[i] = false;
}
// 1 is not a prime number
prime[1] = false;
for(int p = 2; p * p < maxn; p++){
// If prime[p] is not changed, then
// it is a prime
if(prime[p] == true){
// Update all multiples of p
for(int i = p * 2; i < maxn; i += p){
prime[i] = false;
}
}
}
int j = 0;
for(int p = 2; p < maxn; p++) {
if (prime[p]) {
// Storing primes in an array
table[j] = p;
// Update value in primesquare[p * p],
// if p is prime.
if(p < maxn / p) primesquare[p * p] = true;
j++;
}
}
}
// Function to count divisors
int countDivisors(int n)
{
// If number is 1, then it will have only 1
// as a factor. So, total factors will be 1.
if (n == 1)
return 1;
// ans will contain total number of distinct
// divisors
int ans = 1;
// Loop for counting factors of n
for(int i = 0;; i++){
// table[i] is not less than cube root n
if(table[i] * table[i] * table[i] > n)
break;
// Calculating power of table[i] in n.
int cnt = 1; // cnt is power of prime table[i] in n.
while (n % table[i] == 0){ // if table[i] is a factor of n
n = n / table[i];
cnt = cnt + 1; // incrementing power
}
// Calculating the number of divisors
// If n = a^p * b^q then total divisors of n
// are (p+1)*(q+1)
ans = ans * cnt;
}
// if table[i] is greater than cube root of n
// First case
if (prime[n])
ans = ans * 2;
// Second case
else if (primesquare[n])
ans = ans * 3;
// Third case
else if (n != 1)
ans = ans * 4;
return ans; // Total divisors
}
int main()
{
SieveOfEratosthenes();
int sum = 0;
int n = 5;
for(int i = 1; i <= n; i++){
sum += countDivisors(i);
}
cout << sum << endl;
return 0;
}
Output
n = 4 => 8
n = 5 => 10
Complexity
Time complexity: O(n * log(log(n)) + n * n^(1/3))
Space complexity: O(n)
Thanks, #largest_prime_is_463035818 for pointing out my mistake.
So, I started learning C++ recently. This code is trying to add the sum of the squares of each numbers digits. For example: 243: 2*2 + 4*4 + 3*3 = 29.
int sumOfSquareDigits(int n) //BUG WITH INPUT OF 10, 100, 1000, etc.
{
int digits = findDigits(n);
int number;
int remainder;
int *allDigits = new int[digits];
for (int i = 0; i < digits; i++) { //assigns digits to array
if (i + 1 == digits){ //sees if there is a ones value left
allDigits[i] = n;
}
else {
remainder = (n % findPower10(digits - (i + 1)));
number = ((n - remainder) / findPower10(digits - (i + 1)));
allDigits[i] = number; //records leftmost digit
n = n - (allDigits[i] * findPower10(digits - (i + 1))); //gets rid of leftmost number and starts over
}
}
int result = 0;
for (int i = 0; i < digits; i++) { //finds sum of squared digits
result = result + (allDigits[i] * allDigits[i]);
}
delete [] allDigits;
return result;
}
int findDigits(int n) //finds out how many digits the number has
{
int digits = 0;
int test;
do {
digits++;
test = findPower10(digits);
} while (n > test);
return digits;
}
int findPower10(int n) { //function for calculating powers of 10
int result = 1;
for (int i = 0; i < n; i++)
result = result * 10;
return result;
}
And after running the code, I've figured out that it (barely) mostly works. I've found that whenever a user inputs a value of 10, 100, 1000, etc. it always returns a value of 100. I'd like to solve this only using the iostream header.
Sorry if my code isn't too readable or organized! It would also be helpful if there are any shortcuts to my super long code, thanks!
The problem is in the findDigits function. For the values 10, 100, 1000 etc, it calculates the number of the digits minus one. This happens because of the comparison in the loop, you are stopping when n is less or equal to test, but in these cases n is equal test and you should run the next iteration.
So, you should change the line 33:
} while (n > test);
to:
} while (n >= test);
Now, it should work just fine. But it will not work for negative numbers (I don't know this is required, but the solution bellow works for that case too).
I came up with a much simpler solution:
int sumOfSquareDigits(int n)
{
// Variable to mantain the total sum of the squares
int sum = 0;
// This loop will change n until it is zero
while (n != 0) {
/// The current digit we will calculate the square is the rightmost digit,
// so we just get its value using the mod operator
int current = n % 10;
// Add its square to the sum
sum += current*current;
// You divide n by 10, this 'removes' one digit of n
n = n / 10;
}
return sum;
}
I found the problem challenging managed to reduce your code to the following lines:
long long sumOfSquareDigits(long long i) {
long long sum(0L);
do {
long long r = i % 10;
sum += (r * r);
} while(i /= 10);
return sum;
}
Haven't test it thoroughly but I think it works OK.
Given a natural number n (1 <= n <= 500000), please output the summation of all its proper divisors.
Definition: A proper divisor of a natural number is the divisor that is strictly less than the number.
e.g. number 20 has 5 proper divisors: 1, 2, 4, 5, 10, and the divisor summation is: 1 + 2 + 4 + 5 + 10 = 22.
<<--This is a challenge i am trying to do and i am using recursion
int find_sum(std::vector <int> nums,long int sum,int num_now,long int j)
{
if(j<nums[num_now])
{
if(nums[num_now]%j==0)
{
sum=sum+j;
}
return find_sum(nums,sum,num_now,j+1);
}
else
{
return sum;
}
}
sum is the sum of all divisors,nums is the vector i stored number in,num_now is current member in vector,int j is 1 i use it to search for dividers,sadly using this i cant use numbers like 500000 it give's me error,is there any better way to do it or have i made a mistake somewhere.
--Thank you for your time
Here is a recursive way to solve your problem:
int find_sum(int x, int i)
{
if(i == 0)
return 0;
if(x % i == 0)
return i + find_sum(x, (i-1));
return find_sum(x, (i-1));
}
You need to call find_sum(N, N-1); in order to find sum of dividers of N (i must be less than given N because of strict inequality).
In your case it would be find_sum(20, 19);
e.g. my function returns:
71086 for N = 50000
22 for N = 20
0 for N = 1
I don't see the reason why you need to use recursion for solving this problem. I would prefer a more staightforward way to solve it.
long CalculateSumOfDivisors(int number)
{
long sum = 0;
for(int i=1; i<number; i++)
{
// If the remainder of num/i is zero
// then i divides num. So we add it to the
// current sum.
if(number%i==0)
{
sum+=i;
}
}
return sum;
}
Furthermore, we could write a more optimal algorithm, if we note the following:
Let that we have a number n and d is the smallest divisor of n that is greater of 1. (Apparently if the number n is a prime number there is any such a divisor). Then the larget divisor of n is the number n/d.
Based on this we can formulate a more optimal algorithm.
long CalculateSumOfDivisors(int number)
{
int smallestDivisor = FindSmallestDivisor(number);
if(smallestDivisor==1) return 1;
long sum = smallestDivisor;
// Calculate the possible greatest divisor.
int possibleGreatestDivisor = (int)floor(number/smallestDivisor);
for(int i=smallestDivisor+1; i<=possibleGreatestDivisor; i++)
{
if(number%i==0)
{
sum+=i;
}
}
return sum;
}
int FindSmallestDivisor(int number)
{
int smallestDivisor = 1;
for(int i=2; i<number; i++)
{
if(number%i==0)
{
smallestDivisor = i;
break;
}
}
return smallestDivisor;
}
I tried writing code with main function asking user to give the it wants to get sum of. here is the code , hope it helps.
#include<iostream>
using namespace std;
int Sum(int min, int max, int &val, int &sum){
if(min >= max)
return 0;
for ( ; min < max; min++){
if ( val%min == 0){
sum += min + val/min;
return Sum(++min,val/min, val,sum);
}
}
return 0;
}
int main(){
int s=1;
int val;
cout <<"Enter Val to sum:";
cin >> val;
Sum(2,val,val,s);
cout <<"Sum is :"<<s<<endl;
return 0;
}
Here Sum function is used recursively and passed parameters as shown in the code.
Hope it helps.
I don't think you should use recursion.
Instead start by looping from 1..N-1
When you find a divisor adjust the end value for the loop. Example if 2 is a divisor then you know N/2 is also a divisor. And just as important you know there can be no further divisors in the range ]N/2:N[. Likewise if 3 is a divisor then you know N/3 is also a divisor and you know there are no more divisors in the range ]N/3:N[.
Following that concept you can reduced the number of loops significantly for most numbers.
Something like:
long find_sum(int num)
{
long sum = 0;
int max = num;
int i = 1;
while(i < max)
{
if(num % i == 0)
{
sum += i; // Add i to the sum
max = num / i; // Decrement max for performance
if (max != i && max != num)
{
sum += max; // Add max when max isn't equal i
}
}
i++;
}
return sum;
}
Example:
num = 10
sum = 0
i = 1 -> sum = 1, max = 10
i = 2 -> sum = 1+2+5, max = 5
i = 3 -> sum = 1+2+5, max = 5
i = 4 -> sum = 1+2+5, max = 5
i = 5 -> return 8 (1+2+5)
num = 64
sum = 0
i = 1 -> sum = 1, max = 64
i = 2 -> sum = 1+2+32, max = 32
i = 3 -> sum = 1+2+32, max = 32
i = 4 -> sum = 1+2+32+4+16, max = 16
i = 5 -> sum = 1+2+32+4+16, max = 16
i = 6 -> sum = 1+2+32+4+16, max = 16
i = 7 -> sum = 1+2+32+4+16, max = 16
i = 8 -> sum = 1+2+32+4+16+8, max = 8
i = 9 -> return (1+2+32+4+16+8)
The number of loops are kept down by changing max whenever a new divisor is found.
I'm currently learning C++.
I am looking for Hamming numbers (numbers whose prime divisors are less or equal to 5).
When I input a number n, the program should output the n-th Hamming number.
Following numbers are input, and output:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 ...
1 2 3 4 5 6 8 9 10 12 15 16 18 20 24 ...
Finding Hamming numbers looks easy, but increasing the input number increases run time cost exponentially.
If I input over 1000, it almost costs over 1 second,
and over 1200, it almost costs over 5 seconds.
This is the code I wrote:
while (th > 1)
{
h++;
x = h;
while (x % 2 == 0)
x /= 2;
while (x % 3 == 0)
x /= 3;
while (x % 5 == 0)
x /= 5;
if (x == 1)
th--;
}
So I would like to know how I can find the answer faster.
This algorithm doesn't seem to be very good.
Thanks in advance.
Your code is good if you want to check whether one particular number is a hamming number. When you want to build a list of hamming numbers, it is inefficient.
You can use a bottom-up approach: Start with 1 and then recursively multiply that with 2, 3, and 5 to get all hamming numbers up to a certain limit. You have to take care of duplicates, because you can get to 6 by way of 2·3 and 3·2. A set can take care of that.
The code below will generate all hamming numbers that fit into a 32-bit unsigned int. It fills a set by "spreading" to all hamming numbers. Then it constructs a sorted vector from the set, which you can use to find a hamming number at a certain index:
#include <iostream>
#include <algorithm>
#include <set>
#include <vector>
typedef unsigned int uint;
const uint umax = 0xffffffff;
void spread(std::set<uint> &hamming, uint n)
{
if (hamming.find(n) == hamming.end()) {
hamming.insert(n);
if (n < umax / 2) spread(hamming, n * 2);
if (n < umax / 3) spread(hamming, n * 3);
if (n < umax / 5) spread(hamming, n * 5);
}
}
int main()
{
std::set<uint> hamming;
spread(hamming, 1);
std::vector<uint> ordered(hamming.begin(), hamming.end());
for (size_t i = 0; i < ordered.size(); i++) {
std::cout << i << ' ' << ordered[i] << '\n';
}
return 0;
}
This code is faster than your linear method even if you end up creating more hamming numbers than you need.
You don't even need a set if you make sure that you don't construct a number twice. Every hamming number can be written as h = 2^n2 + 3^n3 + 5^n5, so if you find a means to iterate through these uniquely, you're done:
#include <iostream>
#include <algorithm>
#include <set>
#include <vector>
typedef unsigned int uint;
int main()
{
const uint umax = 0xffffffff;
std::vector<uint> hamming;
for (uint k = 1;; k *= 2) {
for (uint l = k;; l *= 3) {
for (uint m = l;; m *= 5) {
hamming.push_back(m);
if (m > umax / 5) break;
}
if (l > umax / 3) break;
}
if (k > umax / 2) break;
}
std::sort(hamming.begin(), hamming.end());
for (size_t i = 0; i < hamming.size(); i++) {
std::cout << i << ' ' << hamming[i] << '\n';
}
return 0;
}
The strange break syntax for the loops is required, because we have to check the size before the overflow. If umax*5 were guananteed not to overflow, these conditions could be written in the condition part of the loop.
The code examples in the Rosetta Code link Koshinae posted use similar strategies, but I'm surprised how lengthy some of them are.
In this link you can find two different solutions for finding the nth hamming number. The second method is the optimized one which can get the result in a few seconds.
/* Function to get the nth ugly number*/
unsigned getNthUglyNo(unsigned n)
{
unsigned ugly[n]; // To store ugly numbers
unsigned i2 = 0, i3 = 0, i5 = 0;
unsigned next_multiple_of_2 = 2;
unsigned next_multiple_of_3 = 3;
unsigned next_multiple_of_5 = 5;
unsigned next_ugly_no = 1;
ugly[0] = 1;
for (int i=1; i<n; i++)
{
next_ugly_no = min(next_multiple_of_2,
min(next_multiple_of_3,
next_multiple_of_5));
ugly[i] = next_ugly_no;
if (next_ugly_no == next_multiple_of_2)
{
i2 = i2+1;
next_multiple_of_2 = ugly[i2]*2;
}
if (next_ugly_no == next_multiple_of_3)
{
i3 = i3+1;
next_multiple_of_3 = ugly[i3]*3;
}
if (next_ugly_no == next_multiple_of_5)
{
i5 = i5+1;
next_multiple_of_5 = ugly[i5]*5;
}
} /*End of for loop (i=1; i<n; i++) */
return next_ugly_no;
}
I am making a random number generator. It asks how many digits the user wants to be in the number. for example it they enter 2 it will generate random numbers between 10 and 99. I have made the generator but my issue is that the numbers are not unique.
Here is my code. I am not sure why it is not generating unique number. I thought srand(time(null)) would do it.
void TargetGen::randomNumberGen()
{
srand (time(NULL));
if (intLength == 1)
{
for (int i = 0; i< intQuantity; i++)
{
int min = 1;
int max = 9;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
else if (intLength == 2)
{
for (int i = 0; i<intQuantity; i++)
{
int min = 10;
int max = 90;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
if (intLength == 3)
{
for (int i = 0; i<intQuantity; i++)
{
int min = 100;
int max = 900;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
else if (intLength == 4)
{
for (int i = 0; i<intQuantity; i++)
{
int min = 1000;
int max = 9000;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
if (intLength == 5)
{
for (int i = 0; i<intQuantity; i++)
{
int min = 10000;
int max = 90000;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
else if (intLength == 6)
{
for (int i = 0; i<intQuantity; i++)
{
int min = 100000;
int max = 900000;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
if (intLength == 7)
{
for (int i = 0; i<intQuantity; i++)
{
int min = 1000000;
int max = 9000000;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
else if (intLength == 8)
{
for (int i = 0; i <intQuantity; i++)
{
int min = 10000000;
int max = 89999999;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
if (intLength == 9)
{
for (int i = 0; i < intQuantity; i++)
{
int min = 100000000;
int max = 900000000;
int number1 = rand();
if (intQuantity > max)
{
intQuantity = max;
}
cout << number1 % max + min << "\t";
}
}
}
Okay so I thought I figured out a way to do this without arrays but It isn't working before I switch to the fisher yates method. Can someone tell me why this isn't working? It is supposed to essentially take the random number put that into variable numGen. Then in variable b = to numgen. Just to hold what numGen used to be so when the loop goes through and generates another random number it will compare it to what the old number is and if it is not equal to it, then it will output it. If it is equal to the old number than rather than outputting it, it will deincrement i so that it will run through the loop without skipping over the number entirely. However, when I do this is infinitely loops. And I am not sure why.
if (intLength == 1)
{
for (int i = 0; i< intQuantity; ++i)
{
int min = 1;
int max = 9;
int number1 = rand();
int numGen = number1 % max + min;
if (intQuantity > max)
{
intQuantity = max;
}
for (int k = 0; k < 1; k++)
{
cout << numGen << "\t";
int b = numGen;
}
int b = numGen;
if (b != numGen )
{
cout << numGen << "\t";
}
else
{
i--;
}
}
}
Everyone has interesting expectations for random numbers -- apparently, you expect random numbers to be unique! If you use any good random number generator, your random numbers will never be guaranteed to be unique.
To make this most obvious, if you wanted to generate random numbers in the range [1, 2], and you were to generate two numbers, you would (normally expect to) get one of the following four possibilities with equal probability:
1, 2
2, 1
1, 1
2, 2
It does not make sense to ask a good random number generator to generate the first two, but not the last two.
Now, take a second to think what to expect if you asked to generate three numbers in the same range... 1, 2, then what??
Uniqueness, therefore, is not, and will not be a property of a random number generator.
Your specific problem may require uniqueness, though. In this case, you need to do some additional work to ensure uniqueness.
One way is to keep a tab on which numbers are already picked. You can keep them in a set, and re-pick if you get one you got earlier. However, this is effective only if you pick a small set of numbers compared to your range; if you pick most of the range, the end of the process gets ineffective.
If the number count you are going to pick corresponds to most of the range, then using an array of the range, and the using a good shuffling algorithm to shuffle the numbers around is a better solution. (The Fisher-Yates shuffle should do the trick.)
Hint 0:
Use Quadratic residue from number theory; an integer q is called a quadratic residue modulo p if it is congruent to a perfect square modulo p; i.e., if there exists an integer x such that:
x2 ≡ q (mod p)
Hint 1:
Theorem: Assuming p is a prime number, the quadratic residue of x is unique as long as 2x < p. For example:
02 ≡ 0 (mod 13)
12 ≡ 1 (mod 13)
22 ≡ 4 (mod 13)
32 ≡ 9 (mod 13)
42 ≡ 3 (mod 13)
52 ≡ 12 (mod 13)
62 ≡ 10 (mod 13)
Hint 2:
Theorem: Assuming p is a prime number such that p ≡ 3 (mod 4), not only x2%p (i.e the quadratic residue) is unique for 2x < p but p - x2%p is also unique for 2x>p. For example:
02%11 = 0
12%11 = 1
22%11 = 4
32%11 = 9
42%11 = 5
52%11 = 3
11 - 62%11 = 8
11 - 72%11 = 6
11 - 82%11 = 2
11 - 92%11 = 7
11 - 102%11 = 10
Thus, this method provides us with a perfect 1-to-1 permutation on the integers less than p, where p can be any prime such that p ≡ 3 (mod 4).
Hint 3:
unsigned int UniqueRandomMapping(unsigned int x)
{
const unsigned int p = 11; //any prime number satisfying p ≡ 3 (mod 4)
unsigned int r = ((unsigned long long) x * x) % p;
if (x <= p / 2) return r;
else return p - r;
}
I didn't worry about the bad input numbers (e.g. out of the range).
Remarks
For 32-bit integers, you may choose the largest prime number such that p ≡ 3 (mod 4) which is less than 232 which is 4294967291.
Even though, this method gives you a 1-to-1 mapping for generating random number, it suffers from the clustering issue.
To improve the randomness of the aforementioned method, combine it with
other unique random mapping methods such as XOR operator.
I'll assume you can come up with a way to figure out how many numbers you want to use. It's pretty simple, since a user input of 2 goes to 10-99, 3 is 100-999, etc.
If you want to come up with your own implementation of unique, randomly generated numbers, check out these links.
http://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle
Here is a very similar implementation: https://stackoverflow.com/a/196065/2142219
In essence, you're creating an array of X integers, all set to the value of their index. You randomly select an index between 0 and MAX, taking the value at this index and swapping it with the max value. MAX is then decremented by 1 and you can repeat it by randomly selecting an index between 0 and MAX - 1.
This gives you a random array of 0-999 integers with no duplicates.
Here are two possible approaches to generating unique random numbers in a range.
Keep track of which numbers you have already generated using std::set, and throw away and regenerate numbers as long as they are already in the set. This approach is not recommended if you want to generate a large number of random numbers, due to the birthday paradox.
Generate all numbers in your given range, take a random permutation of them, and output however many the user wants.
Standard random generators would never generate unique numbers, in this case they would Not be independent.
To generate unique numbers you have to:
Save all number generated and compare new one with old ones, if there is coincidence - regenerate.
or
Use random_shuffle function: http://en.cppreference.com/w/cpp/algorithm/random_shuffle to get all sequence in advance.
Firstly, srand()/rand() commonly have a period of 2^32, which means that after calling srand(), rand() will internally iterate over distinct integers during the first 2^32 calls to rand(). Still, rand() may well return a result with less than 32 bits: such as an int between 0 and RAND_MAX where RAND_MAX is 2^31-1 or 2^15-1, so you may see repeated results as the caller of rand(). You probably read about the period though, or somebody's comment made with awareness of that, and somehow it's been mistaken as uniqueness....
Secondly, given any call to rand() generates a number far larger than you want, and you're doing this...
number1 % max
The result of "number1 % max" is in the range 0 <= N <= max, but the random number itself may have been any multiple of max greater than that. In other words, two distinct random numbers that differ by a multiple of max still produce the same result for number1 % max in your program.
To get distinct random numbers within a range, you could prepopulate a std::vector with all the numbers, then std::shuffle them.