Waste less time on Facebook — follow Brilliant.
×

Factorizing in C++

Can anyone help me out in improving the following program to find factors of number \(n\) as it doesn't work for higher numbers.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
#include<iostream.h>
#include<conio.h>
void main()
{
clrscr();
cin>>n;
for (int a=1;a<n;++a)
{
if (n%a==0)
{
cout<<a<<"\n";
}
}
getch();
}

Note by Pranjal Jain
1 year, 10 months ago

No vote yet
1 vote

Comments

Sort by:

Top Newest

You can firstly improve by running the loop till \( i \leq \frac{n}{2} \). Or you can try this:

1
2
3
4
5
6
7
8
9
int n;
cin>>n;
for(int l=1, int h=n ; l<=h ; l++,h=n/l)
{
    if(n%l==0)
    {
        cout<<l<<" "<<h<<" ";
    }
}
Sudeep Salgia · 1 year, 10 months ago

Log in to reply

@Sudeep Salgia Yes, I noticed \(\frac{n}{2}\) thing. Can you explain algorithm you used? Thanks. Pranjal Jain · 1 year, 10 months ago

Log in to reply

@Pranjal Jain So basically we can pair up the divisors for any given \(n\). So, l will have the smaller one and h will store the larger one ( such that \( l \times h =n \) ) and we are incrementing l by one each time and correspondingly decreasing h. So h will decrease much faster and we would be skipping lot of numbers which anyway we need not check.

PS: I forgot to include the line to check whether \( l= h\) that is if n is a perfect square. Sudeep Salgia · 1 year, 10 months ago

Log in to reply

@Sudeep Salgia It showed many errors. Which compiler are you using? And I believe the condition should be if (n%h==0) Pranjal Jain · 1 year, 10 months ago

Log in to reply

@Sudeep Salgia We can bring down \(\frac{n}{2}\) to \(\lfloor\sqrt{n}\rfloor\). I still can't understand how will it help while dealing with higher numbers. Let me compile. Pranjal Jain · 1 year, 10 months ago

Log in to reply

How many times should i try to help before someone notices Pranjal Jain . I have included a few ideas in the following from Sudeep Salgia as well:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
#include<conio.h>
#include<math.h>
#include<iomanip.h>

int main()
{
double m,i,k=10;
cin>>m;
for(i=1;i<k;i++)
    {
    k=floor(m/i);
    if (i*k==m)
    cout<<setprecision(15)<<i<<"  "<<k<<"  ";
    }
getch();
return 0;
}

Here are the factors of \(10000000099\)

And here are the factors of \(7296872389761\)

Raghav Vaidyanathan · 1 year, 10 months ago

Log in to reply

@Raghav Vaidyanathan Thanks. Sorry for not noticing. I am quite inactive now a days on B'ant. (Turned off all email notifications as well) Pranjal Jain · 1 year, 10 months ago

Log in to reply

Log in to reply

Try long double it has the max and min values \(10^{4932}\quad and \quad 10^{-4932}\) respectively and has a memory of 10 bytes. Its the biggest as far as I know. Harshvardhan Mehta · 1 year, 10 months ago

Log in to reply

@Harshvardhan Mehta Modulus doesn't work with floating data types. Pranjal Jain · 1 year, 10 months ago

Log in to reply

Maybe you could try using long variable or long long . But that won't be really necessary if you are writing this for boards Rohit Shah · 1 year, 10 months ago

Log in to reply

@Rohit Shah I program for fun. -_- I hate boards Pranjal Jain · 1 year, 10 months ago

Log in to reply

@Pranjal Jain Oh ! So have you studied any algorithm books like CLRS ? Or are you active on websites like topcoder , spoj ? Rohit Shah · 1 year, 10 months ago

Log in to reply

@Rohit Shah Currently I am inactive due to JEE. But I will be active after 2-3 months on Project Euler. Pranjal Jain · 1 year, 10 months ago

Log in to reply

Try Java Codes,They are always better than C++ Vraj Mehta · 1 year, 10 months ago

Log in to reply

@Vraj Mehta For that, I'll have to learn java properly. Pranjal Jain · 1 year, 10 months ago

Log in to reply

Try this out

1
2
3
4
5
6
7
8
double n,i;
cin>>n;
for(i=1;i<=(n/2);i++)
    {
    double k=floor(n/i);
    if (i*k==n)
    cout<<setprecision(15)<<i<<endl;
    }
Raghav Vaidyanathan · 1 year, 10 months ago

Log in to reply

@Raghav Vaidyanathan Why are you taking

1
2
Double k=floor(n/i);
    If(i*k==n)

to check if \( i \) is a factor of \( n \) instead of

1
If(n%i == 0)
Siddhartha Srivastava · 1 year, 10 months ago

Log in to reply

@Siddhartha Srivastava Big numbers cannot be stored in long. Therefore we need something which is even bigger, like float or double. The problem with using "%" function with float/double is that it will not work. "%" is meant to work only with integers(int or long or short), It will not work with float and double as they are decimal values. Does 100%2.3 make sense?

Hence we use a workaround to find the true factors of n by using floor. Raghav Vaidyanathan · 1 year, 10 months ago

Log in to reply

@Raghav Vaidyanathan Ouch. I didn't notice that you used double. But then why not use long long? long long has a higher range than a double for the same memory, since a double can only hold 15 significant digits whereas long long can hold upto 18.

Note I don't use C++ so I have no idea if long long is used commonly. Siddhartha Srivastava · 1 year, 10 months ago

Log in to reply

@Siddhartha Srivastava Nope, long long doesn't work in c++. That's the biggest problem here. Raghav Vaidyanathan · 1 year, 10 months ago

Log in to reply

×

Problem Loading...

Note Loading...

Set Loading...