JAVA.MATH.APPROX.E : Approximate e Constant (Java)

Summary

An approximate value of e is used instead of a constant provided by a math library.

Floating point operations are by nature approximated. This can introduce bugs when precise, mathematical properties are expected from inherently imprecise floating point computations. Moreover, computations that might lose precision or overflow, whose result is stored in a larger type, are suspicious since, by computing on the larger type from the beginning, one could avoid approximations and overflows.

Properties

WARNING_FILTER += discard class="Approximate e Constant (Java)"

Example

public class MathConstantApprox
{
    private static double getPI() {
        return 3.14; // Approximate pi Constant (Java) warning issued here
    }

    private static double getE(){
        return 2.71; // Approximate e Constant (Java) warning issued here
    }
    
    private void printCos180(){
        System.out.println("Cos(180°) = " + Math.cos(getPI()));
    }
    
    private  static void printLogE(){
        System.out.println("Log(e) = " + Math.log(getE()));
    }  
    
    public static void main(String[] args){
        printCos180();
        printLogE();
    }
}

Using custom math constants can lead to a loss of precision. For instance, the above program uses custom math constants (PI=3.14 and E=2.71) introducing approximation errors. Indeed, it prints "Cos(180°) = -0,99999873172754" instead of "Cos(180°) = -1" and "Log(e) = 0,99694863489161" instead of "Log(e) = 1".

Instead, use math constants from API libraries, as in the following example.

public class MathConstantApprox
{
    private static double getPI() {
        return Math.PI;
    }

    private static double getE(){
        return Math.E;
    }
    
    private void printCos180(){
        System.out.println("Cos(180°) = " + Math.cos(getPI()));
    }
    
    private  static void printLogE(){
        System.out.println("Log(e) = " + Math.log(getE()));
    }  
    
    public static void main(String[] args){
        printCos180();
        printLogE();
    } 
}

Resolution

Use approximated comparisons (up to some epsilon) rather than exact comparisons. Use integral numbers instead of floating point values, whenever possible. Compute over integral numbers and translate into floating points only at the end, whenever possible. If a large type for the result of a computation is desired, compute from the beginning on that type instead of applying a type conversion at the end of the computation.

Relevant Configuration File Parameters

The following configuration file parameters affect checks for this warning class.

• REPORT_SIMILAR_WARNINGS
• WARNING_FILTER