Numbers that are too large?

[ChesterPlus]

I'm writing a program that displays the odds of winning a lottery. Unfortunately what I wrote is displaying my odds of winning as "1 in inf" rather than an expected numerical output. Is this my IDE or an issue with my code?

#include <iostream>
#include <cstdlib>
#include <ctime>
#include <iomanip>
using namespace std;

// factorial function
float Factorial(int f) {
	float total = 0;
	for (int i = 1; i <= f; i++) {
		total = total * i;
	}
	return total;
}

float Odds(const int nb, const int tr) {
	float sum = 1.0;
	int numballs = nb;
	cout << "Top Range is " << tr << " number balls is " << nb << endl;
	for (int i = tr - nb + 1; i <= tr; i++) {
		sum = (sum * i);
	}
	return sum / Factorial(nb);
}

int main() {
	int num_values = 0;
	int range = 0;
	int num_gens = 0;
	float odds = 0.0;

	// input collection
	cout << "Enter in the number of balls or numbers you wish to pick from.\n";
	while (num_values < 3 || num_values > 7) {
		cout << "Number must be between 3 and 7 inclusive.\n";
		cout << "> ";
		cin >> num_values;
	}
	cout << "\nEnter in the largest number in the lottery.\n";
	while (range < 45 || range > 70) {
		cout << "Number must be between 45 and 70 inclusive.\n";
		cout << "> ";
		cin >> range;
	}
	cout << "\nEnter the number of tickets.\n";
	while (num_gens < 1 || num_gens > 100) {
		cout << "Number must be between 1 and 100 inclusive.\n";
		cout << "> ";
		cin >> num_gens;
	}

	// generate odds
	cout << "\n";
	odds = Odds(num_values, range);

	// output
	cout << "\nYou will select " << num_values << " numbers.\n";
	cout << "The numbers will range from 1 to " << range << ".\n";
	cout << "The odds are 1 in " << odds << ".\n";
}

[salem c]

> total = total * i;
If you initialise total to 0, it doesn't matter how many multiplies you do, it's still 0.

Also, FLT_MAX is typically 3.402823466e+38, whereas 70! = 1.197857167e+100
You need to use a double to represent it.

[lastchance]

You don't need a factorial function at all, so float is OK for a single ticket. (In fact, with those maximum ranges you could do it in a long long int).

You appear to have ignored the number of tickets bought.



For ONE ticket:

float Odds( int nb, int tr )
{
   float result = 1;
   for ( int i = 1; i <= nb; i++) result *= ( tr - i + 1.0 ) / i;
   return result;
}

or you could do it with an integer type if you are careful with the order of multiplication:

unsigned long long Odds( int nb, int tr )
{
   unsigned long long result = 1;
   for ( int i = 1; i <= nb; i++ ) result = result * ( tr - i + 1 ) / i;   // *** BE CAREFUL!
   return result;
}

Later on (when you get around to multiple tickets) you may need to work with probabilities:
1 - (1-p)ⁿ
where p = 1/(odds) and could be very small, so floating-point inaccuracy would suggest the use of a double (at least for p).

[ChesterPlus]

salem c,

I used your fix, thanks much. I initialized my total to 1, and then changed the function to return a double rather than a float, like you said.

All I have to do know is display it as an integer value, and I'm good to go. I can use the <iomanip> library for that I believe.

[jonnin]

iostream to just write an integer. iomanip lets you format your output a bit more. there are many, many ways to format and handle output, string stream, direct string manipulation (eg std::tostring), and the new format tools (https://en.cppreference.com/w/cpp/utility/format) to name a few.

generally I prefer a premade lookup table of things like factorial where it takes a loop to find the next value but you only can fit a small # of values in standard data types.