algorithm - a lot of decimal value storage in C++ -

i wanted write program returns how many months survive when we're given our monthly expenditure, amount of disposable income, , interest rate (all in integers). instance, if start disposable income = 1000, interest rate = 5%, monthly expenditure = 100,

after first month: 1000*1.05 - 100 = 950, have 950 dollars left after second month: = 950*1.05 - 100 = 897.5  

and on, , in end, can survive 14 months.

i wrote following code in c++:

int main(){  int i=0; int initialvalue; int interest; int monthly;  double value=1.0*initialvalue; double r=1+1.0*interest/100; while(value > 0){     if(value < monthly){         break;     }     else     {     value=value*r-monthly;       i++;     } }; cout<<i; return 0; } 

but sufficiently large values of initialvalue , small values of monthly, program wrote runs degree it's unusable. there problem code makes run not (or slow)?

any appreciated.

double cannot store numbers precisely. 1 consequence of when subtract small number large number, result not changed original large value.

one solution problem use int calculations. think of values number of pennies, rather number of dollars.