Counting unneeded part in double datatype

Question

I have a weird problem with datatype:double.

Here is my code:

public class Example {

    /**
     * @param args
     */
    public static void main(String[] args) {
        // TODO Auto-generated method stub
        double a = 1.0;
        for(int i =0; i<10;i++){
            System.out.println("Number => " + a );
            a += 0.1;
        }
    }

}

Output should be :

Number => 1.0
Number => 1.1
Number => 1.2
Number => 1.3
Number => 1.4
Number => 1.5
Number => 1.6
Number => 1.7
Number => 1.8
Number => 1.9

But, this code sample's result is :

Number => 1.0
Number => 1.1
Number => 1.2000000000000002
Number => 1.3000000000000003
Number => 1.4000000000000004
Number => 1.5000000000000004
Number => 1.6000000000000005
Number => 1.7000000000000006
Number => 1.8000000000000007
Number => 1.9000000000000008

I use eclipse to compile this code block. I tried it on netbeans but nothing changed.

How should it be happen ? Any ideas?

Regards

If exact representation of short decimal fractions matters to you, you should be using BigDecimal, not double. Double is based on binary fractions, and gives no special importance to ten or its powers. It is a more general way of storing and manipulating a useful approximation to a wide range of numbers. — Patricia Shanahan, May 30 '13 at 14:18
I am looking for this for a long time. Almost everyone say that bigDecimal is the solution. But i just wonder why it happens like this. Btw, my que is dup, i did not noticed the other question. — alicanbatur, May 30 '13 at 14:21

Evgeniy Dorofeev · Answer 1 · 2013-05-30T14:23:33.917

1

It's not so easy. I would use Commons-Math

for (int i = 0; i < 10; i++) {
    a = Precision.round(a, 1);   // <- round to 1 digit after decimal point
    ...

I think the only Java SE alternative to that is

a = Double.parseDouble(String.format(Locale.US, "%.1f", a));

edited May 30 '13 at 14:23

answered May 30 '13 at 14:14

Evgeniy Dorofeev

133,369
30
199
275

Thanks for your response. It is quite useful when i use small numbers. But i am working on a project which is about calculating money. – alicanbatur May 30 '13 at 14:32
@alicanbatur Do not use `Double` for that. Use `BigDecimal`, see http://stackoverflow.com/questions/285680/representing-monetary-values-in-java – Sotirios Delimanolis May 30 '13 at 14:33
Then use BigDecimal, primitive double is no good for money – Evgeniy Dorofeev May 30 '13 at 14:34
Yes @SotiriosDelimanolis i am using BigDecimal by now. Thanks for all of your advices. – alicanbatur May 30 '13 at 14:34

MaheshVarma · Answer 2 · 2013-05-30T15:18:33.253

0

According to the javadoc

If at least one of the operands to a numerical operator is of type double, then the operation is carried out using 64-bit floating-point arithmetic, and the result of the numerical operator is a value of type double. If the other operand is not a double, it is first widened (§5.1.5) to type double by numeric promotion (§5.6).

Here is the Source

edited May 30 '13 at 15:18

answered May 30 '13 at 14:26

MaheshVarma

2,081
7
35
58

Counting unneeded part in double datatype

2 Answers2