How to nicely format floating numbers to String without unnecessary decimal 0?

https://stackoverflow.com/questions/703396

22-08-2019
|

Question

An 64-bit double can represent integer +/- 2⁵³ exactly

Given this fact I choose to use a double type as a single type for all my types, since my largest integer is unsigned 32-bit.

But now I have to print these pseudo integers, but the problem is they are also mixed in with actual doubles.

So how do I print these doubles nicely in Java?

I have tried String.format("%f", value), which is close, except I get a lot of trailing zeros for small values.

Here's an example output of of %f

232.00000000
0.18000000000
1237875192.0
4.5800000000
0.00000000
1.23450000

What I want is:

Sure I can write a function to trim those zeros, but that's lot of performance loss due to String manipulation. Can I do better with another format code?

EDIT

The answers by Tom E. and Jeremy S. are unacceptable as they both arbitrarily rounds to 2 decimal places. Please understand the problem before answering.

EDIT 2

Please note that String.format(format, args...) is locale-dependent (see answers below).

Solution

If the idea is to print integers stored as doubles as if they are integers, and otherwise print the doubles with the minimum necessary precision:

public static String fmt(double d)
{
    if(d == (long) d)
        return String.format("%d",(long)d);
    else
        return String.format("%s",d);
}

Produces:

And does not rely on string manipulation.

OTHER TIPS

new DecimalFormat("#.##").format(1.199); //"1.2"

As pointed in the comments, this is not the right answer to the original question.
That said, it is a very useful way to format numbers without unnecessary trailing zeros.

String.format("%.2f", value) ;

In short:

If you want to get rid of trailing zeros and Locale problems, then you should use :

double myValue = 0.00000021d;

DecimalFormat df = new DecimalFormat("0", DecimalFormatSymbols.getInstance(Locale.ENGLISH));
df.setMaximumFractionDigits(340); //340 = DecimalFormat.DOUBLE_FRACTION_DIGITS

System.out.println(df.format(myValue)); //output: 0.00000021

Explanation:

Why other answers did not suit me :

Double.toString() or System.out.println or FloatingDecimal.toJavaFormatString uses scientific notations if double is less than 10^-3 or greater than or equal to 10^7
```
double myValue = 0.00000021d;
String.format("%s", myvalue); //output: 2.1E-7
```
by using %f, the default decimal precision is 6, otherwise you can hardcode it but it results in extra zeros added if you have less decimals. Example :
```
double myValue = 0.00000021d;
String.format("%.12f", myvalue); //output: 0.000000210000
```

by using setMaximumFractionDigits(0); or %.0f you remove any decimal precision, which is fine for integers/longs but not for double

double myValue = 0.00000021d;
System.out.println(String.format("%.0f", myvalue)); //output: 0
DecimalFormat df = new DecimalFormat("0");
System.out.println(df.format(myValue)); //output: 0

by using DecimalFormat, you are local dependent. In French locale, the decimal separator is a comma, not a point :
```
double myValue = 0.00000021d;
DecimalFormat df = new DecimalFormat("0");
df.setMaximumFractionDigits(340);
System.out.println(df.format(myvalue));//output: 0,00000021
```
Using the ENGLISH locale makes sure you get a point for decimal separator, wherever your program will run

Why using 340 then for setMaximumFractionDigits ?

Two reasons :

setMaximumFractionDigits accepts an integer but its implementation has a maximum digits allowed of DecimalFormat.DOUBLE_FRACTION_DIGITS which equals 340
Double.MIN_VALUE = 4.9E-324 so with 340 digits you are sure not to round your double and loose precision

Why not:

if (d % 1.0 != 0)
    return String.format("%s", d);
else
    return String.format("%.0f",d);

This should work with the extreme values supported by Double. Yields:

My 2 cents:

if(n % 1 == 0) {
    return String.format(Locale.US, "%.0f", n));
} else {
    return String.format(Locale.US, "%.1f", n));
}

On my machine, the following function is roughly 7 times faster than the function provided by JasonD's answer, since it avoids String.format:

public static String prettyPrint(double d) {
  int i = (int) d;
  return d == i ? String.valueOf(i) : String.valueOf(d);
}

Naw, never mind.

Performance loss due to String manipulation is zero.

And here's the code to trim the end after %f

private static String trimTrailingZeros(String number) {
    if(!number.contains(".")) {
        return number;
    }

    return number.replaceAll("\\.?0*$", "");
}

if (d == Math.floor(d)) {
    return String.format("%.0f", d);
} else {
    return Double.toString(d);
}

Please note that String.format(format, args...) is locale-dependent because it formats using the user's default locale, that is, probably with commas and even spaces inside like 123 456,789 or 123,456.789, which may be not exactly what you expect.

You may prefer to use String.format((Locale)null, format, args...).

For example,

    double f = 123456.789d;
    System.out.println(String.format(Locale.FRANCE,"%f",f));
    System.out.println(String.format(Locale.GERMANY,"%f",f));
    System.out.println(String.format(Locale.US,"%f",f));

prints

123456,789000
123456,789000
123456.789000

and this is what will String.format(format, args...) do in different countries.

EDIT Ok, since there has been a discussion about formalities:

    res += stripFpZeroes(String.format((Locale) null, (nDigits!=0 ? "%."+nDigits+"f" : "%f"), value));
    ...

protected static String stripFpZeroes(String fpnumber) {
    int n = fpnumber.indexOf('.');
    if (n == -1) {
        return fpnumber;
    }
    if (n < 2) {
        n = 2;
    }
    String s = fpnumber;
    while (s.length() > n && s.endsWith("0")) {
        s = s.substring(0, s.length()-1);
    }
    return s;
}

I made a DoubleFormatter to efficiently convert a great numbers of double values to a nice/presentable String:

double horribleNumber = 3598945.141658554548844; 
DoubleFormatter df = new DoubleFormatter(4,6); //4 = MaxInteger, 6 = MaxDecimal
String beautyDisplay = df.format(horribleNumber);

If the integer part of V has more than MaxInteger => display V in scientist format (1.2345e+30) otherwise display in normal format 124.45678.
the MaxDecimal decide numbers of decimal digits (trim with banker's rounding)

Here the code:

import java.math.RoundingMode;
import java.text.DecimalFormat;
import java.text.DecimalFormatSymbols;
import java.text.NumberFormat;
import java.util.Locale;

import com.google.common.base.Preconditions;
import com.google.common.base.Strings;

/**
 * Convert a double to a beautiful String (US-local):
 * 
 * double horribleNumber = 3598945.141658554548844; 
 * DoubleFormatter df = new DoubleFormatter(4,6);
 * String beautyDisplay = df.format(horribleNumber);
 * String beautyLabel = df.formatHtml(horribleNumber);
 * 
 * Manipulate 3 instances of NumberFormat to efficiently format a great number of double values.
 * (avoid to create an object NumberFormat each call of format()).
 * 
 * 3 instances of NumberFormat will be reused to format a value v:
 * 
 * if v < EXP_DOWN, uses nfBelow
 * if EXP_DOWN <= v <= EXP_UP, uses nfNormal
 * if EXP_UP < v, uses nfAbove
 * 
 * nfBelow, nfNormal and nfAbove will be generated base on the precision_ parameter.
 * 
 * @author: DUONG Phu-Hiep
 */
public class DoubleFormatter
{
    private static final double EXP_DOWN = 1.e-3;
    private double EXP_UP; // always = 10^maxInteger
    private int maxInteger_;
    private int maxFraction_;
    private NumberFormat nfBelow_; 
    private NumberFormat nfNormal_;
    private NumberFormat nfAbove_;

    private enum NumberFormatKind {Below, Normal, Above}

    public DoubleFormatter(int maxInteger, int maxFraction){
        setPrecision(maxInteger, maxFraction);
    }

    public void setPrecision(int maxInteger, int maxFraction){
        Preconditions.checkArgument(maxFraction>=0);
        Preconditions.checkArgument(maxInteger>0 && maxInteger<17);

        if (maxFraction == maxFraction_ && maxInteger_ == maxInteger) {
            return;
        }

        maxFraction_ = maxFraction;
        maxInteger_ = maxInteger;
        EXP_UP =  Math.pow(10, maxInteger);
        nfBelow_ = createNumberFormat(NumberFormatKind.Below);
        nfNormal_ = createNumberFormat(NumberFormatKind.Normal);
        nfAbove_ = createNumberFormat(NumberFormatKind.Above);
    }

    private NumberFormat createNumberFormat(NumberFormatKind kind) {
        final String sharpByPrecision = Strings.repeat("#", maxFraction_); //if you do not use Guava library, replace with createSharp(precision);
        NumberFormat f = NumberFormat.getInstance(Locale.US);

        //Apply banker's rounding:  this is the rounding mode that statistically minimizes cumulative error when applied repeatedly over a sequence of calculations
        f.setRoundingMode(RoundingMode.HALF_EVEN);

        if (f instanceof DecimalFormat) {
            DecimalFormat df = (DecimalFormat) f;
            DecimalFormatSymbols dfs = df.getDecimalFormatSymbols();

            //set group separator to space instead of comma

            //dfs.setGroupingSeparator(' ');

            //set Exponent symbol to minus 'e' instead of 'E'
            if (kind == NumberFormatKind.Above) {
                dfs.setExponentSeparator("e+"); //force to display the positive sign in the exponent part
            } else {
                dfs.setExponentSeparator("e");
            }

            df.setDecimalFormatSymbols(dfs);

            //use exponent format if v is out side of [EXP_DOWN,EXP_UP]

            if (kind == NumberFormatKind.Normal) {
                if (maxFraction_ == 0) {
                    df.applyPattern("#,##0");
                } else {
                    df.applyPattern("#,##0."+sharpByPrecision);
                }
            } else {
                if (maxFraction_ == 0) {
                    df.applyPattern("0E0");
                } else {
                    df.applyPattern("0."+sharpByPrecision+"E0");
                }
            }
        }
        return f;
    } 

    public String format(double v) {
        if (Double.isNaN(v)) {
            return "-";
        }
        if (v==0) {
            return "0"; 
        }
        final double absv = Math.abs(v);

        if (absv<EXP_DOWN) {
            return nfBelow_.format(v);
        }

        if (absv>EXP_UP) {
            return nfAbove_.format(v);
        }

        return nfNormal_.format(v);
    }

    /**
     * format and higlight the important part (integer part & exponent part) 
     */
    public String formatHtml(double v) {
        if (Double.isNaN(v)) {
            return "-";
        }
        return htmlize(format(v));
    }

    /**
     * This is the base alogrithm: create a instance of NumberFormat for the value, then format it. It should
     * not be used to format a great numbers of value 
     * 
     * We will never use this methode, it is here only to understanding the Algo principal:
     * 
     * format v to string. precision_ is numbers of digits after decimal. 
     * if EXP_DOWN <= abs(v) <= EXP_UP, display the normal format: 124.45678
     * otherwise display scientist format with: 1.2345e+30 
     * 
     * pre-condition: precision >= 1
     */
    @Deprecated
    public String formatInefficient(double v) {

        final String sharpByPrecision = Strings.repeat("#", maxFraction_); //if you do not use Guava library, replace with createSharp(precision);

        final double absv = Math.abs(v);

        NumberFormat f = NumberFormat.getInstance(Locale.US);

        //Apply banker's rounding:  this is the rounding mode that statistically minimizes cumulative error when applied repeatedly over a sequence of calculations
        f.setRoundingMode(RoundingMode.HALF_EVEN);

        if (f instanceof DecimalFormat) {
            DecimalFormat df = (DecimalFormat) f;
            DecimalFormatSymbols dfs = df.getDecimalFormatSymbols();

            //set group separator to space instead of comma

            dfs.setGroupingSeparator(' ');

            //set Exponent symbol to minus 'e' instead of 'E'

            if (absv>EXP_UP) {
                dfs.setExponentSeparator("e+"); //force to display the positive sign in the exponent part
            } else {
                dfs.setExponentSeparator("e");
            }
            df.setDecimalFormatSymbols(dfs);

            //use exponent format if v is out side of [EXP_DOWN,EXP_UP]

            if (absv<EXP_DOWN || absv>EXP_UP) {
                df.applyPattern("0."+sharpByPrecision+"E0");
            } else {
                df.applyPattern("#,##0."+sharpByPrecision);
            }
        }
        return f.format(v);
    }

    /**
     * Convert "3.1416e+12" to "<b>3</b>.1416e<b>+12</b>"
     * It is a html format of a number which highlight the integer and exponent part
     */
    private static String htmlize(String s) {
        StringBuilder resu = new StringBuilder("<b>");
        int p1 = s.indexOf('.');

        if (p1>0) {
            resu.append(s.substring(0, p1));
            resu.append("</b>");
        } else {
            p1 = 0;
        }

        int p2 = s.lastIndexOf('e');
        if (p2>0) {
            resu.append(s.substring(p1, p2));
            resu.append("<b>");
            resu.append(s.substring(p2, s.length()));
            resu.append("</b>");
        } else {
            resu.append(s.substring(p1, s.length()));
            if (p1==0){
                resu.append("</b>");
            }
        }
        return resu.toString();
    }
}

Note: I used 2 functions from GUAVA library. If you don't use GUAVA, code it yourself:

/**
 * Equivalent to Strings.repeat("#", n) of the Guava library: 
 */
private static String createSharp(int n) {
    StringBuilder sb = new StringBuilder(); 
    for (int i=0;i<n;i++) {
        sb.append('#');
    }
    return sb.toString();
}

Use a DecimalFormat and setMinimumFractionDigits(0)

This one will get the job done nicely, I know the topic is old, but I was struggling with the same issue till I came to this. I hope someone find it useful.

    public static String removeZero(double number) {
        DecimalFormat format = new DecimalFormat("#.###########");
        return format.format(number);
    }

new DecimalFormat("00.#").format(20.236)
//out =20.2

new DecimalFormat("00.#").format(2.236)
//out =02.2

0 for minimum number of digits
Renders # digits

String s = String.valueof("your int variable");
while (g.endsWith("0") && g.contains(".")) {
    g = g.substring(0, g.length() - 1);
    if (g.endsWith("."))
    {
        g = g.substring(0, g.length() - 1);
    }
}

Late answer but...

You said you choose to store your numbers with the double type. I think this could be the root of the problem because it forces you to store integers into doubles (and therefore losing the initial information about the value's nature). What about storing your numbers in instances of the Number class (superclass of both Double and Integer) and rely on polymorphism to determine the correct format of each number ?

I know it may not be acceptable to refactor a whole part of your code due to that but it could produce the desired output without extra code/casting/parsing.

Example:

import java.util.ArrayList;
import java.util.List;

public class UseMixedNumbers {

    public static void main(String[] args) {
        List<Number> listNumbers = new ArrayList<Number>();

        listNumbers.add(232);
        listNumbers.add(0.18);
        listNumbers.add(1237875192);
        listNumbers.add(4.58);
        listNumbers.add(0);
        listNumbers.add(1.2345);

        for (Number number : listNumbers) {
            System.out.println(number);
        }
    }

}

Will produce the following output:

public static String fmt(double d) {
    String val = Double.toString(d);
    String[] valArray = val.split("\\.");
    long valLong = 0;
    if(valArray.length == 2){
        valLong = Long.parseLong(valArray[1]);
    }
    if (valLong == 0)
        return String.format("%d", (long) d);
    else
        return String.format("%s", d);
}

I had to use this cause d == (long)d was giving me violation in sonar report

This is what I came up with:

  private static String format(final double dbl) {
    return dbl % 1 != 0 ? String.valueOf(dbl) : String.valueOf((int) dbl);
  }

Simple one liner, only casts to int if really need to

Here are two ways to achieve it. First, the shorter (and probably better) way:

public static String formatFloatToString(final float f)
  {
  final int i=(int)f;
  if(f==i)
    return Integer.toString(i);
  return Float.toString(f);
  }

And here's the longer and probably worse way:

public static String formatFloatToString(final float f)
  {
  final String s=Float.toString(f);
  int dotPos=-1;
  for(int i=0;i<s.length();++i)
    if(s.charAt(i)=='.')
      {
      dotPos=i;
      break;
      }
  if(dotPos==-1)
    return s;
  int end=dotPos;
  for(int i=dotPos+1;i<s.length();++i)
    {
    final char c=s.charAt(i);
    if(c!='0')
      end=i+1;
    }
  final String result=s.substring(0,end);
  return result;
  }

Here is an answer that actually works (combination of different answers here)

public static String removeTrailingZeros(double f)
{
    if(f == (int)f) {
        return String.format("%d", (int)f);
    }
    return String.format("%f", f).replaceAll("0*$", "");
}

I know this is a really old thread.. But I think the best way to do this is as below:

public class Test {

    public static void main(String args[]){
        System.out.println(String.format("%s something",new Double(3.456)));
        System.out.println(String.format("%s something",new Double(3.456234523452)));
        System.out.println(String.format("%s something",new Double(3.45)));
        System.out.println(String.format("%s something",new Double(3)));
    }
}

Output:

3.456 something
3.456234523452 something
3.45 something
3.0 something

The only issue is the last one where .0 doesn't get removed. But if you are able to live with that then this works best. %.2f will round it to the last 2 decimal digits. So will DecimalFormat. If you need all the decimal places but not the trailing zeros then this works best.

String s = "1.210000";
while (s.endsWith("0")){
    s = (s.substring(0, s.length() - 1));
}

This will make the string to drop the tailing 0-s.

Licensed under: CC-BY-SA with attribution

Not affiliated with StackOverflow