How can I only show three digists of the double value

Question

I have double values like 0.0345333 or 1.0897777 and I want to convert them to 0.034 and 1.089 in Java?

How can I do it?

I want to convert the values, not print them since I need it for calculation.

I find the following way but I am looking for simple way. double score=(tff* idf); DecimalFormat df = new DecimalFormat("#.###"); p.score=Double.parseDouble(df.format(score));

@m0skit0 I did see those, never a bad thing to ask however. Its so easy to not notice — Richard Tingle, May 05 '14 at 15:00
This is just for outputting or for using this values for other calculations? — m0skit0, May 05 '14 at 15:01
@RichardTingle Then let's ask if he's sure if it's Java, or sure if it's a programming language he's talking about... ;) — m0skit0, May 05 '14 at 15:01
@m0skit0 *sigh* you'll notice that a whole load of answers and duplicate votes assuming rounding rather than truncating. I'm not sure why you think its so terrible me encouraging making the distiction clear — Richard Tingle, May 05 '14 at 15:03
You need to use the `BigDecimal`. Have a look at thia - [Double vs. BigDecimal?](http://stackoverflow.com/questions/3413448/double-vs-bigdecimal). — Bhesh Gurung, May 05 '14 at 15:04
I wrote the following code but I want easeir way. double score=(tff* idf); DecimalFormat df = new DecimalFormat("#.###"); p.score=Double.parseDouble(df.format(score)); — user3586943, May 05 '14 at 15:04
If it's for calculation then you need to use the `BigDecimal`. You can achieve that with `double`. `BigDecimal` is like the WYSIWYG version of `double`. — Bhesh Gurung, May 05 '14 at 15:10

Gogutz · Answer 1 · 2014-05-05T15:11:29.550

2

If you want to display the result use.

String.format("%1$,.3f", myDouble);

But if you need to truncate the numbers you could use:

double number = 1.0897777;
double result = ((int)(number*1000))/1000d;//1.089

edited May 05 '14 at 15:11

answered May 05 '14 at 15:04

Gogutz

2,005
17
19

This prints the value, it doesn't convert it. – Anubian Noob May 05 '14 at 15:06
Is there a way I only work with double not string. – user3586943 May 05 '14 at 15:07
@AnubianNoob I see you are quite quick at giving minuses... Well, thanks for that... i guess. – Gogutz May 05 '14 at 15:12
@Gogutz What? Also this now answers the question. – Anubian Noob May 05 '14 at 15:13
The second answer you give is the best answer to the question asked--I don't think it's a great solution because of the way doubles work which is why Scale is not in the Double class but is in the BigDecimal class. – Bill K May 05 '14 at 16:18

Anubian Noob · Accepted Answer · 2014-05-05T16:11:22.117

1

This will truncate value to 3 decimal places:

value = ((int)(value*1000))/1000.0;

This will round to 3 decimal places:

value = (double)Math.round(value*1000)/1000;

For example:

double d1 = 0.0345333;
d1 = (double)Math.round(d1*1000)/1000;
System.out.println(d1); // 0.035

double d2 = 0.0345333;
d2 = ((int)(d2 *1000))/1000.0;
System.out.println(d2); // 0.034

edited May 05 '14 at 16:11

answered May 05 '14 at 15:01

Anubian Noob

13,426
6
53
75

It is not work it changes value to 0 and 1 respectivley – user3586943 May 05 '14 at 15:03
If the value is an integer it won't work, if it's a double it will. – Anubian Noob May 05 '14 at 15:04
The value are double but it is not work for me – user3586943 May 05 '14 at 15:06
I think `Math.round` returns a `long` so you want to be dividing by `1000.0`. – 2rs2ts May 05 '14 at 15:10
@2rs2ts That's why I casted to double. – Anubian Noob May 05 '14 at 15:11
@AnubianNoob ah, sorry, I thought the case would apply after the division. I tried it on ideone and it worked (although it gave me `0.035`.) – 2rs2ts May 05 '14 at 15:12
This rounds, not truncates, to 3 decimal places `value = Math.round(value*1000)/1000.0;` – Peter Lawrey May 05 '14 at 15:40

score 1 · Answer 3 · answered May 05 '14 at 15:19

1

You can also use BigDecimal

BigDecimal value = new BigDecimal(1.0897777);
value = value.setScale(3, RoundingMode.DOWN);

answered May 05 '14 at 15:19

d.moncada

16,900
5
53
82

1

If you are doing "Rounding" type calculations and want to continue tow work with the results rather than just display them, then this is almost ceratinly the best answer. A double or float will almost always get jacked up during such a rounding operation (rounding 1.0897777 back into a double will end up as 1.089000001 or something). – Bill K May 05 '14 at 16:16

How can I only show three digists of the double value

3 Answers3