I've got a problem with rounding in JavaScript. I'm using a function for rounding:
function roundup(rnum, rlength){
var newnumber = Math.round(rnum * Math.pow(10, rlength)) / Math.pow(10, rlength);
return newnumber;
}
var amount = roundup(2253.825, 3);
Strange thing is, when I round up the number 2253.825, the result is 2253.82 which has to be 2253.83. When I round up the number 5592.825 the result is 5592.83, which is correct.
Any idea how to fix this?
Floating point rounding errors are at fault here. 2620.825 * 100 is 262082.49999999997, so when you round it, the result is 262082.
Here is a more robust approach that corrects for this:
function roundup(rnum, rlength) {
var shifted = rnum * Math.pow(10, rlength),
rounded = Math.round(shifted),
delta = Math.abs(shifted - rounded);
if (delta > 0.4999999 && delta < 0.5) {
rounded += (rounded < 0 ? -1 : 1);
}
return rounded / Math.pow(10, rlength);
}
console.log("2620.825 :=> " + roundup(2620.825, 2));
console.log("2621.825 :=> " + roundup(2621.825, 2));
console.log("2620.8255 :=> " + roundup(2620.8255, 2));
The problem shows up inherently because the way IEEE 754 doubles work. If you really want to always get mathematically correct results, you need to use a custom datatype, for example a BigDecimal implementation like this one.
Please also take a look at this answer, as it may be useful.
Related
in JavaScript, the typical way to round a number to N decimal places is something like:
function roundNumber(num, dec) {
return Math.round(num * Math.pow(10, dec)) / Math.pow(10, dec);
}
function roundNumber(num, dec) {
return Math.round(num * Math.pow(10, dec)) / Math.pow(10, dec);
}
console.log(roundNumber(0.1 + 0.2, 2));
console.log(roundNumber(2.1234, 2));
However this approach will round to a maximum of N decimal places while I want to always round to N decimal places. For example "2.0" would be rounded to "2".
Any ideas?
I think that there is a more simple approach to all given here, and is the method Number.toFixed() already implemented in JavaScript.
simply write:
var myNumber = 2;
myNumber.toFixed(2); //returns "2.00"
myNumber.toFixed(1); //returns "2.0"
etc...
I found a way. This is Christoph's code with a fix:
function toFixed(value, precision) {
var precision = precision || 0,
power = Math.pow(10, precision),
absValue = Math.abs(Math.round(value * power)),
result = (value < 0 ? '-' : '') + String(Math.floor(absValue / power));
if (precision > 0) {
var fraction = String(absValue % power),
padding = new Array(Math.max(precision - fraction.length, 0) + 1).join('0');
result += '.' + padding + fraction;
}
return result;
}
Read the details of repeating a character using an array constructor here if you are curious as to why I added the "+ 1".
That's not a rounding ploblem, that is a display problem. A number doesn't contain information about significant digits; the value 2 is the same as 2.0000000000000. It's when you turn the rounded value into a string that you have make it display a certain number of digits.
You could just add zeroes after the number, something like:
var s = number.toString();
if (s.indexOf('.') == -1) s += '.';
while (s.length < s.indexOf('.') + 4) s += '0';
(Note that this assumes that the regional settings of the client uses period as decimal separator, the code needs some more work to function for other settings.)
There's always a better way for doing things. Use toPrecision -
var number = 51.93999999999761;
I would like to get four digits precision: 51.94
just do:
number.toPrecision(4);
the result will be: 51.94
This works for rounding to N digits (if you just want to truncate to N digits remove the Math.round call and use the Math.trunc one):
function roundN(value, digits) {
var tenToN = 10 ** digits;
return /*Math.trunc*/(Math.round(value * tenToN)) / tenToN;
}
Had to resort to such logic at Java in the past when I was authoring data manipulation E-Slate components. That is since I had found out that adding 0.1 many times to 0 you'd end up with some unexpectedly long decimal part (this is due to floating point arithmetics).
A user comment at Format number to always show 2 decimal places calls this technique scaling.
Some mention there are cases that don't round as expected and at http://www.jacklmoore.com/notes/rounding-in-javascript/ this is suggested instead:
function round(value, decimals) {
return Number(Math.round(value+'e'+decimals)+'e-'+decimals);
}
PHP-Like rounding Method
The code below can be used to add your own version of Math.round to your own namespace which takes a precision parameter. Unlike Decimal rounding in the example above, this performs no conversion to and from strings, and the precision parameter works same way as PHP and Excel whereby a positive 1 would round to 1 decimal place and -1 would round to the tens.
var myNamespace = {};
myNamespace.round = function(number, precision) {
var factor = Math.pow(10, precision);
var tempNumber = number * factor;
var roundedTempNumber = Math.round(tempNumber);
return roundedTempNumber / factor;
};
myNamespace.round(1234.5678, 1); // 1234.6
myNamespace.round(1234.5678, -1); // 1230
from Mozilla Developer reference for Math.round()
Hopefully working code (didn't do much testing):
function toFixed(value, precision) {
var precision = precision || 0,
neg = value < 0,
power = Math.pow(10, precision),
value = Math.round(value * power),
integral = String((neg ? Math.ceil : Math.floor)(value / power)),
fraction = String((neg ? -value : value) % power),
padding = new Array(Math.max(precision - fraction.length, 0) + 1).join('0');
return precision ? integral + '.' + padding + fraction : integral;
}
I think below function can help
function roundOff(value,round) {
return (parseInt(value * (10 ** (round + 1))) - parseInt(value * (10 ** round)) * 10) > 4 ? (((parseFloat(parseInt((value + parseFloat(1 / (10 ** round))) * (10 ** round))))) / (10 ** round)) : (parseFloat(parseInt(value * (10 ** round))) / ( 10 ** round));
}
usage : roundOff(600.23458,2); will return 600.23
function roundton(num, n) {
return Number(num.toFixed(n));
}
This uses JS's built-in method Number.prototype.toFixed which is meant for formatting strings but allows us to round to a specific number of digits. the Number() call converts it back to a number object cleanly
Ideally, we wouldn't need to convert it to a string, but toFixed is written in native C++ doing basic cstring operations so it's likely still fast.
If you do not really care about rounding, just added a toFixed(x) and then removing trailing 0es and the dot if necessary. It is not a fast solution.
function format(value, decimals) {
if (value) {
value = value.toFixed(decimals);
} else {
value = "0";
}
if (value.indexOf(".") < 0) { value += "."; }
var dotIdx = value.indexOf(".");
while (value.length - dotIdx <= decimals) { value += "0"; } // add 0's
return value;
}
I used below two methods :
Number.prototype.myRound = function (decimalPlaces) {
var multiplier = Math.pow(10, decimalPlaces);
return (Math.round(this * multiplier) / multiplier);
};
alert((239.525).myRound(2));
Mathematically alert should be 239.53 but its giving 239.52 as output.
So i tried using .toFixed() function & i got proper answer.
But when i try to get answer for 239.575 it gives again wrong output.
alert((239.575).toFixed(2));
Here output should be 239.58 instead its giving 239.57.
This error creating a bit difference in final output. So can anyone help me to sort this out?
This method will give very correct round result.
function RoundNum(num, length) {
var number = Math.round(num * Math.pow(10, length)) / Math.pow(10, length);
return number;
}
Just call this method.
alert(RoundNum(192.168,2));
Internally, 239.575 cannot be represented exactly. In binary, 0.575 would be something like 1/2 + 1/16 + 1/128 + 1/256 + ....
It just so happens that, represented in binary, the result is slightly less than 239.575. Therefore, Math.round rounds down.
To demonstrate, try this:
alert(239.575 - 239.5)
You would expect the result to be 0.075, but instead you get 0.07499999999998863.
Just use Math.round
function round(figureToRound){
var roundOff = Math.round((figureToRound* 100 ).toFixed(2))/100;
return roundOff;
}
console.log(round(1.005));
This will help the rounding off issue completely.
round() will do the trick.Try This:
var v= Math.round(239.575 * 100) / 100;
alert(v);
Working FIddle
The problem is probably floating point inaccuracy, thus you might get different results in different cases (different gathering of a number, different browsers etc.).
See also this: toFixed(2) rounds "x.525" inconsistently?
In my software I use this:
(require DecimalJS)
Number.prototype.toFixed = function(fixed) {
return (new Decimal(Number(this))).toFixed(parseFloat(fixed) ||
0);
};
var x = 1.005;
console.log( x.toFixed(2) ); //1.01
function bestRound(val, decimals){
decimals = decimals || 2;
var multiplier = Math.pow(10, decimals)
return Math.round((val * multiplier ).toFixed(decimals)) / multiplier;
}
bestRound(239.575 - 239.5) 0.08
bestRound(239.575) 239.58
bestRound(239.525) 239.53
bestRound(1.005) 1.01
I got this to simply overwrite it ->
Number.prototype.toFixed = function(fractionDigits, returnAsString = true) {
var digits = parseInt(fractionDigits) || 0;
var num = Number(this);
if( isNaN(num) ) {
return 'NaN';
}
var sign = num < 0 ? -1 : 1;
if (sign < 0) { num = -num; }
digits = Math.pow(10, digits);
num *= digits;
//num = Math.round(num.toFixed(12));
num = Math.round( Math.round(num * Math.pow(10,12)) / Math.pow(10,12) );
var ret = sign * num / digits;
return (returnAsString ? ret.toString() : ret ); // tofixed returns as string always
}
This question already has answers here:
How to round to at most 2 decimal places, if necessary
(91 answers)
Closed 5 years ago.
I have the following JavaScript syntax:
var discount = Math.round(100 - (price / listprice) * 100);
This rounds up to the whole number. How can I return the result with two decimal places?
NOTE - See Edit 4 if 3 digit precision is important
var discount = (price / listprice).toFixed(2);
toFixed will round up or down for you depending on the values beyond 2 decimals.
Example: http://jsfiddle.net/calder12/tv9HY/
Documentation: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Number/toFixed
Edit - As mentioned by others this converts the result to a string. To avoid this:
var discount = +((price / listprice).toFixed(2));
Edit 2- As also mentioned in the comments this function fails in some precision, in the case of 1.005 for example it will return 1.00 instead of 1.01. If accuracy to this degree is important I've found this answer: https://stackoverflow.com/a/32605063/1726511 Which seems to work well with all the tests I've tried.
There is one minor modification required though, the function in the answer linked above returns whole numbers when it rounds to one, so for example 99.004 will return 99 instead of 99.00 which isn't ideal for displaying prices.
Edit 3 - Seems having the toFixed on the actual return was STILL screwing up some numbers, this final edit appears to work. Geez so many reworks!
var discount = roundTo((price / listprice), 2);
function roundTo(n, digits) {
if (digits === undefined) {
digits = 0;
}
var multiplicator = Math.pow(10, digits);
n = parseFloat((n * multiplicator).toFixed(11));
var test =(Math.round(n) / multiplicator);
return +(test.toFixed(digits));
}
See Fiddle example here: https://jsfiddle.net/calder12/3Lbhfy5s/
Edit 4 - You guys are killing me. Edit 3 fails on negative numbers, without digging into why it's just easier to deal with turning a negative number positive before doing the rounding, then turning it back before returning the result.
function roundTo(n, digits) {
var negative = false;
if (digits === undefined) {
digits = 0;
}
if (n < 0) {
negative = true;
n = n * -1;
}
var multiplicator = Math.pow(10, digits);
n = parseFloat((n * multiplicator).toFixed(11));
n = (Math.round(n) / multiplicator).toFixed(digits);
if (negative) {
n = (n * -1).toFixed(digits);
}
return n;
}
Fiddle: https://jsfiddle.net/3Lbhfy5s/79/
If you use a unary plus to convert a string to a number as documented on MDN.
For example:+discount.toFixed(2)
The functions Math.round() and .toFixed() is meant to round to the nearest integer. You'll get incorrect results when dealing with decimals and using the "multiply and divide" method for Math.round() or parameter for .toFixed(). For example, if you try to round 1.005 using Math.round(1.005 * 100) / 100 then you'll get the result of 1, and 1.00 using .toFixed(2) instead of getting the correct answer of 1.01.
You can use following to solve this issue:
Number(Math.round(100 - (price / listprice) * 100 + 'e2') + 'e-2');
Add .toFixed(2) to get the two decimal places you wanted.
Number(Math.round(100 - (price / listprice) * 100 + 'e2') + 'e-2').toFixed(2);
You could make a function that will handle the rounding for you:
function round(value, decimals) {
return Number(Math.round(value + 'e' + decimals) + 'e-' + decimals);
}
Example:
https://jsfiddle.net/k5tpq3pd/36/
Alternative
You can add a round function to Number using prototype. I would not suggest adding .toFixed() here as it would return a string instead of number.
Number.prototype.round = function(decimals) {
return Number((Math.round(this + "e" + decimals) + "e-" + decimals));
}
and use it like this:
var numberToRound = 100 - (price / listprice) * 100;
numberToRound.round(2);
numberToRound.round(2).toFixed(2); //Converts it to string with two decimals
Example
https://jsfiddle.net/k5tpq3pd/35/
Source: http://www.jacklmoore.com/notes/rounding-in-javascript/
To get the result with two decimals, you can do like this :
var discount = Math.round((100 - (price / listprice) * 100) * 100) / 100;
The value to be rounded is multiplied by 100 to keep the first two digits, then we divide by 100 to get the actual result.
The best and simple solution I found is
function round(value, decimals) {
return Number(Math.round(value+'e'+decimals)+'e-'+decimals);
}
round(1.005, 2); // 1.01
try using discount.toFixed(2);
I think the best way I've seen it done is multiplying by 10 to the power of the number of digits, then doing a Math.round, then finally dividing by 10 to the power of digits. Here is a simple function I use in typescript:
function roundToXDigits(value: number, digits: number) {
value = value * Math.pow(10, digits);
value = Math.round(value);
value = value / Math.pow(10, digits);
return value;
}
Or plain javascript:
function roundToXDigits(value, digits) {
if(!digits){
digits = 2;
}
value = value * Math.pow(10, digits);
value = Math.round(value);
value = value / Math.pow(10, digits);
return value;
}
A small variation on the accepted answer.
toFixed(2) returns a string, and you will always get two decimal places. These might be zeros. If you would like to suppress final zero(s), simply do this:
var discount = + ((price / listprice).toFixed(2));
Edited:
I've just discovered what seems to be a bug in Firefox 35.0.1, which means that the above may give NaN with some values.
I've changed my code to
var discount = Math.round(price / listprice * 100) / 100;
This gives a number with up to two decimal places. If you wanted three, you would multiply and divide by 1000, and so on.
The OP wants two decimal places always, but if toFixed() is broken in Firefox it needs fixing first.
See https://bugzilla.mozilla.org/show_bug.cgi?id=1134388
Fastest Way - faster than toFixed():
TWO DECIMALS
x = .123456
result = Math.round(x * 100) / 100 // result .12
THREE DECIMALS
x = .123456
result = Math.round(x * 1000) / 1000 // result .123
function round(num,dec)
{
num = Math.round(num+'e'+dec)
return Number(num+'e-'+dec)
}
//Round to a decimal of your choosing:
round(1.3453,2)
Here is a working example
var value=200.2365455;
result=Math.round(value*100)/100 //result will be 200.24
To handle rounding to any number of decimal places, a function with 2 lines of code will suffice for most needs. Here's some sample code to play with.
var testNum = 134.9567654;
var decPl = 2;
var testRes = roundDec(testNum,decPl);
alert (testNum + ' rounded to ' + decPl + ' decimal places is ' + testRes);
function roundDec(nbr,dec_places){
var mult = Math.pow(10,dec_places);
return Math.round(nbr * mult) / mult;
}
I knew javascript could have rounding issue with divisions, but not with multiplication. How do you solve those?
var p = $('input[name="productsUS"]').val().replace(",", ".");
var t = $('input[name="productsWorld"]').val().replace(",", ".");
if (p >= 0 && t >= 1) {
var r = p / t;
r = Math.round(r * 10000) / 10000;
var aff = (r * 100) + "%";
if p = 100 and t = 57674
r = 0.0017 (ok) and aff = 0.16999999999999998% (arg)
How could I obtain aff = 0.17?
("0.16999999999999998").tofixed(2) gives you 0.17.
var aff = (r * 100).toFixed(2) + "%";
Live DEMO
toFixed on MDN
If you want to aff to remain a Number instead of being converted to a String, you can use toFixed to work around the precision issues and then "cast" it back to a number using the unary + operator like so:
var n = 0.16999999999999998;
n = +n.toFixed(10); // 0.17
You probably want to use a higher precision than 2 decimal places to avoid rounding issues, I used 10 here.
That's a bug in JavaScript (although it also affects a few other languages, such as Python), due to the way it stores non-whole numbers being a bit buggy (binary!). The only way to work around it is to round it to two decimal places.
in JavaScript, the typical way to round a number to N decimal places is something like:
function roundNumber(num, dec) {
return Math.round(num * Math.pow(10, dec)) / Math.pow(10, dec);
}
function roundNumber(num, dec) {
return Math.round(num * Math.pow(10, dec)) / Math.pow(10, dec);
}
console.log(roundNumber(0.1 + 0.2, 2));
console.log(roundNumber(2.1234, 2));
However this approach will round to a maximum of N decimal places while I want to always round to N decimal places. For example "2.0" would be rounded to "2".
Any ideas?
I think that there is a more simple approach to all given here, and is the method Number.toFixed() already implemented in JavaScript.
simply write:
var myNumber = 2;
myNumber.toFixed(2); //returns "2.00"
myNumber.toFixed(1); //returns "2.0"
etc...
I found a way. This is Christoph's code with a fix:
function toFixed(value, precision) {
var precision = precision || 0,
power = Math.pow(10, precision),
absValue = Math.abs(Math.round(value * power)),
result = (value < 0 ? '-' : '') + String(Math.floor(absValue / power));
if (precision > 0) {
var fraction = String(absValue % power),
padding = new Array(Math.max(precision - fraction.length, 0) + 1).join('0');
result += '.' + padding + fraction;
}
return result;
}
Read the details of repeating a character using an array constructor here if you are curious as to why I added the "+ 1".
That's not a rounding ploblem, that is a display problem. A number doesn't contain information about significant digits; the value 2 is the same as 2.0000000000000. It's when you turn the rounded value into a string that you have make it display a certain number of digits.
You could just add zeroes after the number, something like:
var s = number.toString();
if (s.indexOf('.') == -1) s += '.';
while (s.length < s.indexOf('.') + 4) s += '0';
(Note that this assumes that the regional settings of the client uses period as decimal separator, the code needs some more work to function for other settings.)
There's always a better way for doing things. Use toPrecision -
var number = 51.93999999999761;
I would like to get four digits precision: 51.94
just do:
number.toPrecision(4);
the result will be: 51.94
This works for rounding to N digits (if you just want to truncate to N digits remove the Math.round call and use the Math.trunc one):
function roundN(value, digits) {
var tenToN = 10 ** digits;
return /*Math.trunc*/(Math.round(value * tenToN)) / tenToN;
}
Had to resort to such logic at Java in the past when I was authoring data manipulation E-Slate components. That is since I had found out that adding 0.1 many times to 0 you'd end up with some unexpectedly long decimal part (this is due to floating point arithmetics).
A user comment at Format number to always show 2 decimal places calls this technique scaling.
Some mention there are cases that don't round as expected and at http://www.jacklmoore.com/notes/rounding-in-javascript/ this is suggested instead:
function round(value, decimals) {
return Number(Math.round(value+'e'+decimals)+'e-'+decimals);
}
PHP-Like rounding Method
The code below can be used to add your own version of Math.round to your own namespace which takes a precision parameter. Unlike Decimal rounding in the example above, this performs no conversion to and from strings, and the precision parameter works same way as PHP and Excel whereby a positive 1 would round to 1 decimal place and -1 would round to the tens.
var myNamespace = {};
myNamespace.round = function(number, precision) {
var factor = Math.pow(10, precision);
var tempNumber = number * factor;
var roundedTempNumber = Math.round(tempNumber);
return roundedTempNumber / factor;
};
myNamespace.round(1234.5678, 1); // 1234.6
myNamespace.round(1234.5678, -1); // 1230
from Mozilla Developer reference for Math.round()
Hopefully working code (didn't do much testing):
function toFixed(value, precision) {
var precision = precision || 0,
neg = value < 0,
power = Math.pow(10, precision),
value = Math.round(value * power),
integral = String((neg ? Math.ceil : Math.floor)(value / power)),
fraction = String((neg ? -value : value) % power),
padding = new Array(Math.max(precision - fraction.length, 0) + 1).join('0');
return precision ? integral + '.' + padding + fraction : integral;
}
I think below function can help
function roundOff(value,round) {
return (parseInt(value * (10 ** (round + 1))) - parseInt(value * (10 ** round)) * 10) > 4 ? (((parseFloat(parseInt((value + parseFloat(1 / (10 ** round))) * (10 ** round))))) / (10 ** round)) : (parseFloat(parseInt(value * (10 ** round))) / ( 10 ** round));
}
usage : roundOff(600.23458,2); will return 600.23
function roundton(num, n) {
return Number(num.toFixed(n));
}
This uses JS's built-in method Number.prototype.toFixed which is meant for formatting strings but allows us to round to a specific number of digits. the Number() call converts it back to a number object cleanly
Ideally, we wouldn't need to convert it to a string, but toFixed is written in native C++ doing basic cstring operations so it's likely still fast.
If you do not really care about rounding, just added a toFixed(x) and then removing trailing 0es and the dot if necessary. It is not a fast solution.
function format(value, decimals) {
if (value) {
value = value.toFixed(decimals);
} else {
value = "0";
}
if (value.indexOf(".") < 0) { value += "."; }
var dotIdx = value.indexOf(".");
while (value.length - dotIdx <= decimals) { value += "0"; } // add 0's
return value;
}