Lost, non-functioning if statements in my first program

[DanStirletz]

The program delivers all the values of the expression and the loop correctly. However the if statements do not work. What is wrong with my code and how can I fix it?

CODE:
using namespace std;
int main()


{
double x, y;

cout << "My name is Daniil Aleksandrov, and this is the output of my first program" <<endl <<endl;
cout << "X Value\t\tY Value" << endl <<endl <<endl;


for (x = -4; x<=2.0; x++){

y = (15 * x * x * x - (4 * x +54))/(sqrt(5 * x * x + 1) + 7 * fabs(x - 2.5));

cout << "X is " << x << "\tY is " << y;

if(y = 0)
cout << "\tY is Zero";
cout << endl;

if(y > 0)
cout << "\tY is Positive";
cout << endl;

if(y < 0)
cout << "\tY is Negative";
cout << endl;


}

cout << "\nMy First Program is Complete" <<endl;

return 0;
}

OUTPUT:

My name is Daniil Alek, and this is the output of my first program

X Value Y Value


X is -4 Y is -18.3119


X is -3 Y is -9.8714


X is -2 Y is -4.60056


X is -1 Y is -2.41192


X is 0 Y is -2.91892


X is 1 Y is -3.32059


X is 2 Y is 7.17593



My First Program is Complete

Process returned 0 (0x0) execution time : 0.031 s
Press any key to continue.

[TheIdeasMan]

if(y = 0)//assignment

should be:

if(y == 0.0) //equality - almost certain to be false

Always use the equality operator in an if statement.

But that is a problem in itself, FP (floating point - includes doubles and floats) numbers are represented as binary fractions, and cannot represent all real numbers. If you specify x as a double number what you get is x +/- a small number (about 1e-16).

For the same reason you should not do this:

for (x = -4; x<=2.0; x++){ // probably executes 8 times not 7

Always use integer values for the conditionals in loops - you can always convert them to double inside the loop.

To make meaningful comparisons with FP numbers, you need the absolute value of a the number with an arbitrary precision (0.001 say). So for comparison with zero you have:

double Precision = 0.001;
if (abs(x) < Precision) //close enough to zero for us 

http://www.cplusplus.com/reference/cmath/abs/

There is also a value called epsilon which you can use instead of the arbitrary precision.

http://www.cplusplus.com/reference/limits/numeric_limits/

#include <limits>       // std::numeric_limits
double Precision = std::numeric_limits<double>::epsilon()

if (abs(x) < Precision) {}

Hope this helps & good luck :)

[DanStirletz]

Thank you so much. That cleared it up.