Develop Reference

Is it sufficient to convert a double to a BigDecimal just before addition to retain original precision?

We are solving a numeric precision related bug. Our system collects some numbers and spits their sum.
The issue is that the system does not retain the numeric precision, e.g. 300.7 + 400.9 = 701.599..., while expected result would be 701.6. The precision is supposed to adapt to the input values so we cannot just round results to fixed precision.
The problem is obvious, we use double for the values and addition accumulates the error from the binary representation of the decimal value.
The path of the data is following:
XML file, type xsd:decimal
Parse into a java primitive double. Its 15 decimal places should be enough, we expect values no longer than 10 digits total, 5 fraction digits.
Store into DB MySql 5.5, type double
Load via Hibernate into a JPA entity, i.e. still primitive double
Sum bunch of these values
Print the sum into another XML file
Now, I assume the optimal solution would be converting everything to a decimal format. Unsurprisingly, there is a pressure to go with the cheapest solution. It turns out that converting doubles to BigDecimal just before adding a couple of numbers works in case B in following example:
import java.math.BigDecimal;
public class Arithmetic {
public static void main(String[] args) {
double a = 0.3;
double b = -0.2;
// A
System.out.println(a + b);//0.09999999999999998
// B
System.out.println(BigDecimal.valueOf(a).add(BigDecimal.valueOf(b)));//0.1
// C
System.out.println(new BigDecimal(a).add(new BigDecimal(b)));//0.099999999999999977795539507496869191527366638183593750
}
}
More about this:
Why do we need to convert the double into a string, before we can convert it into a BigDecimal?
Unpredictability of the BigDecimal(double) constructor
I am worried that such a workaround would be a ticking bomb.
First, I am not so sure that this arithmetic is bullet proof for all cases.
Second, there is still some risk that someone in the future might implement some changes and change B to C, because this pitfall is far from obvious and even a unit test may fail to reveal the bug.
I would be willing to live with the second point but the question is: Would this workaround provide correct results? Could there be a case where somehow
Double.valueOf("12345.12345").toString().equals("12345.12345")
is false? Given that Double.toString, according to javadoc, prints just the digits needed to uniquely represent underlying double value, so when parsed again, it gives the same double value? Isn't that sufficient for this use case where I only need to add the numbers and print the sum with this magical Double.toString(Double d) method? To be clear, I do prefer what I consider the clean solution, using BigDecimal everywhere, but I am kind of short of arguments to sell it, by which I mean ideally an example where conversion to BigDecimal before addition fails to do the job described above.

If you can't avoid parsing into primitive double or store as double, you should convert to BigDecimal as early as possible.
double can't exactly represent decimal fractions. The value in double x = 7.3; will never be exactly 7.3, but something very very close to it, with a difference visible from the 16th digit or so on to the right (giving 50 decimal places or so). Don't be mislead by the fact that printing might give exactly "7.3", as printing already does some kind of rounding and doesn't show the number exactly.
If you do lots of computations with double numbers, the tiny differences will eventually sum up until they exceed your tolerance. So using doubles in computations where decimal fractions are needed, is indeed a ticking bomb.
[...] we expect values no longer than 10 digits total, 5 fraction digits.
I read that assertion to mean that all numbers you deal with, are to be exact multiples of 0.00001, without any further digits. You can convert doubles to such BigDecimals with
new BigDecimal.valueOf(Math.round(doubleVal * 100000), 5)
This will give you an exact representation of a number with 5 decimal fraction digits, the 5-fraction-digits one that's closest to the input doubleVal. This way you correct for the tiny differences between the doubleVal and the decimal number that you originally meant.
If you'd simply use BigDecimal.valueOf(double val), you'd go through the string representation of the double you're using, which can't guarantee that it's what you want. It depends on a rounding process inside the Double class which tries to represent the double-approximation of 7.3 (being maybe 7.30000000000000123456789123456789125) with the most plausible number of decimal digits. It happens to result in "7.3" (and, kudos to the developers, quite often matches the "expected" string) and not "7.300000000000001" or "7.3000000000000012" which both seem equally plausible to me.
That's why I recommend not to rely on that rounding, but to do the rounding yourself by decimal shifting 5 places, then rounding to the nearest long, and constructing a BigDecimal scaled back by 5 decimal places. This guarantees that you get an exact value with (at most) 5 fractional decimal places.
Then do your computations with the BigDecimals (using the appropriate MathContext for rounding, if necessary).
When you finally have to store the number as a double, use BigDecimal.doubleValue(). The resulting double will be close enough to the decimal that the above-mentioned conversion will surely give you the same BigDecimal that you had before (unless you have really huge numbers like 10 digits before the decimal point - the you're lost with double anyway).
P.S. Be sure to use BigDecimal only if decimal fractions are relevant to you - there were times when the British Shilling currency consisted of twelve Pence. Representing fractional Pounds as BigDecimal would give a disaster much worse than using doubles.

It depends on the Database you are using. If you are using SQL Server you can use data type as numeric(12, 8) where 12 represent numeric value and 8 represents precision. similarly, for my SQL DECIMAL(5,2) you can use.
You won't lose any precision value if you use the above-mentioned datatype.
Java Hibernate Class :
You can define
private double latitude;
Database:

Java BigDecimal is rounding necessary with addition?

I am working with BigDecimal and I know that if I divide I have to use MathContext and tell which Scale and RoundingMode to avoid ArithmeticException as described in the documentation:
In the case of divide, the exact quotient could have an infinitely
long decimal expansion; for example, 1 divided by 3. If the quotient
has a nonterminating decimal expansion and the operation is specified
to return an exact result, an ArithmeticException is thrown.
In the method I'm working on I have to sum amounts coming from our database (that are rounded at 2 decimals) with amounts coming from an external service, and I don't know the exact scaling of these amounts (probabily 3 decimals).
My question is, can I trust BigDecimal's add method and use it without rounding and scaling or it's a good practice to always specify the desired scale?
Is there any particular case in which addition and subtraction can raise ArithmeticException?

BigDecimal.add() will throw ArithmeticException if the scale of the result does not fit into an int.
A simple example is adding two numbers with the maximum and the minimum scales:
BigDecimal a = new BigDecimal(BigInteger.ONE, Integer.MIN_VALUE);
BigDecimal b = new BigDecimal(BigInteger.ONE, Integer.MAX_VALUE);
a.add(b);
If your application needs to operate at such scale then you probably have some bigger problems than worrying about arithmetic exception.
Adding numbers without using MathContext will maintain proper scale and give you the precice result. Depending on the actual values this approach can use arbitrary amount of memory to represent increasingly long numbers, and longer numbers take more time to add.
Adding numbers without using MathContext and rouding once after the summation will give you the precice result rounded to the requested MathContext. The memory and computation costs are the same as in the first case.
Using MathContext for each addition will produce a result which can differ from the presize result by an arbitrary value, but the memory and speed will be more predicatable.
Choosing which one of these approaches to use really depends on the nature of the task, so it is up to you to assess and chose the proper approach for each particular case.

Floating point types returned in ORM / DSL

The Java™ Tutorials state that "this data type [double] should never be used for precise values, such as currency." Is the fact that an ORM / DSL is returning floating point numbers for database columns storing values to be used to calculate monetary amounts a problem? I'm using QueryDSL and I'm dealing with money. QueryDSL is returning a Double for any number with a precision up to 16 and a BigDecimal thereafter. This concerns me as I'm aware that floating point arithmetic isn't suitable for currency calculations.
From this QueryDSL issue I'm led to believe that Hibernate does the same thing; see OracleDialect. Why does it use a Double rather than a BigDecimal? Is it safe to retrieve the Double and construct a BigDecimal, or is there a chance that a number with a precision of less than 16 could be incorrectly represented? Is it only when performing arithmetic operations that a Double can have floating-point issues, or are there values to which it cannot be accurately initialised?

Using floating point numbers for storing money is a bad idea indeed. Floating points can approximate an operation result, but that's not what you want when dealing with money.
The easiest way to fix it, in a database portable way, is to simply store cents. This is the proffered way of dealing with currency operations in financial operations. Pay attention that most databases use the half-away from zero rounding algorithm, so make sure that's appropriate in your context.
When it comes to money you should always ask a local accountant, especially for the rounding part. Better safe then sorry.
Now back to your questions:
Is it safe to retrieve the Double and construct a BigDecimal, or is
there a chance that a number with a precision of less than 16 could be
incorrectly represented?
This is a safe operation as long as your database uses at most a 16 digit precision. If it uses a higher precision, you'd need to override the OracleDialect and
Is it only when performing arithmetic operations that a Double can
have floating-point issues, or are there values to which it cannot be
accurately initialised?
When performing arithmetic operations you must always take into consideration the monetary rounding anyway, and that applies to BigDecimal as well. So if you can guarantee that the database value doesn't loose any decimal when being cast to a java Double, you are fine to create a BigDecimal from it. Using BigDecimal pays off when applying arithmetic operations to the database loaded value.
As for the threshold of 16, according to Wiki:
The 11 bit width of the exponent allows the representation of numbers
with a decimal exponent between 10−308 and 10308, with full 15–17
decimal digits precision. By compromising precision, subnormal
representation allows values smaller than 10−323.

There seems to be several concerns mentioned in the question, comments, and answers by Robert Bain. I've collected and paraphrased some of these.
Is it safe to use a double to store a precise value?
Yes, provided the number of significant-digits (precision) is small enough.
From wikipedia
If a decimal string with at most 15 significant digits is converted to IEEE 754 double precision representation and then converted back to a string with the same number of significant digits, then the final string should match the original.
But new BigDecimal(1000.1d) has the value 1000.1000000000000227373675443232059478759765625, why not 1000.1?
In the quote above I added emphasis - when converted from a double the number of significant digits must be specified, e.g.
new BigDecimal(1000.1d, new MathContext(15))
Is it safe to use a double for arbitrary arithmetic on precise values?
No, each intermediate value used in the calculation could introduce additional error.
Using a double to store exact values should be seen as an optimization. It introduces risk that if care is not taken, precision could be lost. Using a BigDecimal is much less likely to have unexpected consequences and should be your default choice.
Is it correct that QueryDSL returns a double for precise value?
It is not necessarily incorrect, but is probably not desirable. I would suggest you engage with the QueryDSL developers... but I see you have already raised an issue and they intend to change this behavior.

After much deliberation, I must conclude that the answer to my own question:
Is the fact that an ORM / DSL is returning floating point numbers for database columns storing values to be used to calculate monetary amounts a problem?
put simply, is yes. Please read on.
Is it safe to retrieve the Double and construct a BigDecimal, or is there a chance that a number with a precision of less than 16 could be incorrectly represented?
A number with a precision of less than 16 decimal digits is incorrectly represented in the following example.
BigDecimal foo = new BigDecimal(1000.1d);
The BigDecimal value of foo is 1000.1000000000000227373675443232059478759765625. 1000.1 has a precision of 1 and is being misrepresented from precision 14 of the BigDecimal value.
Is it only when performing arithmetic operations that a Double can have floating-point issues, or are there values to which it cannot be accurately initialised?
As per the example above, there are values to which it cannot be accurately initialised. As The Java™ Tutorials clearly states, "This data type [float / double] should never be used for precise values, such as currency. For that, you will need to use the java.math.BigDecimal class instead."
Interestingly, calling BigDecimal.valueOf(someDouble) appeared at first to magically resolve things but upon realising that it calls Double.toString() then reading Double's documentation it became apparent that this is not appropriate for exact values either.
In conclusion, when dealing with exact values, floating point numbers are never appropriate. As such, in my mind, ORMs / DSLs should be mapping to BigDecimal unless otherwise specified, given that most database use will involve the calculation of exact values.
Update:
Based on this conclusion, I've raised this issue with QueryDSL.

It is not only about arithmetic operations, but also about pure read&write.
Oracle NUMBER and BigDecimal do both use decadic base. So when you read number from database and then you store it back you can be sure, that the same number was written. (Unless it exceeds Oracle's limit of 38 digits).
If you convert NUMBER into binary base (Double) and then you convert it back do decadic then you might expect problems. And also this operation must be much slower.

Wrong Output Dollar Amount To Coins [duplicate]

double r = 11.631;
double theta = 21.4;
In the debugger, these are shown as 11.631000000000000 and 21.399999618530273.
How can I avoid this?

These accuracy problems are due to the internal representation of floating point numbers and there's not much you can do to avoid it.
By the way, printing these values at run-time often still leads to the correct results, at least using modern C++ compilers. For most operations, this isn't much of an issue.

I liked Joel's explanation, which deals with a similar binary floating point precision issue in Excel 2007:
See how there's a lot of 0110 0110 0110 there at the end? That's because 0.1 has no exact representation in binary... it's a repeating binary number. It's sort of like how 1/3 has no representation in decimal. 1/3 is 0.33333333 and you have to keep writing 3's forever. If you lose patience, you get something inexact.
So you can imagine how, in decimal, if you tried to do 3*1/3, and you didn't have time to write 3's forever, the result you would get would be 0.99999999, not 1, and people would get angry with you for being wrong.

If you have a value like:
double theta = 21.4;
And you want to do:
if (theta == 21.4)
{
}
You have to be a bit clever, you will need to check if the value of theta is really close to 21.4, but not necessarily that value.
if (fabs(theta - 21.4) <= 1e-6)
{
}

This is partly platform-specific - and we don't know what platform you're using.
It's also partly a case of knowing what you actually want to see. The debugger is showing you - to some extent, anyway - the precise value stored in your variable. In my article on binary floating point numbers in .NET, there's a C# class which lets you see the absolutely exact number stored in a double. The online version isn't working at the moment - I'll try to put one up on another site.
Given that the debugger sees the "actual" value, it's got to make a judgement call about what to display - it could show you the value rounded to a few decimal places, or a more precise value. Some debuggers do a better job than others at reading developers' minds, but it's a fundamental problem with binary floating point numbers.

Use the fixed-point decimal type if you want stability at the limits of precision. There are overheads, and you must explicitly cast if you wish to convert to floating point. If you do convert to floating point you will reintroduce the instabilities that seem to bother you.
Alternately you can get over it and learn to work with the limited precision of floating point arithmetic. For example you can use rounding to make values converge, or you can use epsilon comparisons to describe a tolerance. "Epsilon" is a constant you set up that defines a tolerance. For example, you may choose to regard two values as being equal if they are within 0.0001 of each other.
It occurs to me that you could use operator overloading to make epsilon comparisons transparent. That would be very cool.
For mantissa-exponent representations EPSILON must be computed to remain within the representable precision. For a number N, Epsilon = N / 10E+14
System.Double.Epsilon is the smallest representable positive value for the Double type. It is too small for our purpose. Read Microsoft's advice on equality testing

I've come across this before (on my blog) - I think the surprise tends to be that the 'irrational' numbers are different.
By 'irrational' here I'm just referring to the fact that they can't be accurately represented in this format. Real irrational numbers (like π - pi) can't be accurately represented at all.
Most people are familiar with 1/3 not working in decimal: 0.3333333333333...
The odd thing is that 1.1 doesn't work in floats. People expect decimal values to work in floating point numbers because of how they think of them:
1.1 is 11 x 10^-1
When actually they're in base-2
1.1 is 154811237190861 x 2^-47
You can't avoid it, you just have to get used to the fact that some floats are 'irrational', in the same way that 1/3 is.

One way you can avoid this is to use a library that uses an alternative method of representing decimal numbers, such as BCD

If you are using Java and you need accuracy, use the BigDecimal class for floating point calculations. It is slower but safer.

Seems to me that 21.399999618530273 is the single precision (float) representation of 21.4. Looks like the debugger is casting down from double to float somewhere.

You cant avoid this as you're using floating point numbers with fixed quantity of bytes. There's simply no isomorphism possible between real numbers and its limited notation.
But most of the time you can simply ignore it. 21.4==21.4 would still be true because it is still the same numbers with the same error. But 21.4f==21.4 may not be true because the error for float and double are different.
If you need fixed precision, perhaps you should try fixed point numbers. Or even integers. I for example often use int(1000*x) for passing to debug pager.

Dangers of computer arithmetic

If it bothers you, you can customize the way some values are displayed during debug. Use it with care :-)
Enhancing Debugging with the Debugger Display Attributes

Refer to General Decimal Arithmetic
Also take note when comparing floats, see this answer for more information.

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

Is BigDecimal an overkill in some cases?

I'm working with money so I need my results to be accurate but I only need a precision of 2 decimal points (cents). Is BigDecimal needed to guarantee results of multiplication/division are accurate?

BigDecimal is a very appropriate type for decimal fraction arithmetic with a known number of digits after the decimal point. You can use an integer type and keep track of the multiplier yourself, but that involves doing in your code work that could be automated.
As well as managing the digits after the decimal point, BigDecimal will also expand the number of stored digits as needed - many business and government financial calculations involve sums too large to store in cents in an int.
I would consider avoiding it only if you need to store a very large array of amounts of money, and are short of memory.

One common option is to do all your calculation with integer or long(the cents value) and then simply add two decimal places when you need to display it.
Similarly, there is a JODA Money library that will give you a more full-featured API for money calculations.

It depends on your application. One reason to use that level of accuracy is to prevent errors accumulated over many operations from percolating up and causing loss of valuable information. If you're creating a casual application and/or are only using it for, say, data entry, BigDecimal is very likely overkill.

+1 for Patricias answer, but I very strongly discourage anyone to implement own classes with an integer datatype with fixed bitlength as long as someone really do not know what you are doing. BigDecimal supports all rounding and precision issues while a long/int has severe problems:
Unknown number of fraction digits: Trade exchanges/Law/Commerce are varying in their amount
of fractional digits, so you do not know if your chosen number of digits must be changed and
adjusted in the future. Worse: There are some things like stock evaluation which need a ridiculous amount of fractional digits. A ship with 1000 metric tons of coal causes e.g.
4,12 € costs of ice, leading to 0,000412 €/ton.
Unimplemented operations: It means that people are likely to use floating-point for
rounding/division or other arithmetic operations, hiding the inexactness and leading to
all the known problems of floating-point arithmetic.
Overflow/Underflow: After reaching the maximum amount, adding an amount results in changing the sign. Long.MAX_VALUE switches to Long.MIN_VALUE. This can easily happen if you are doing fractions like (a*b*c*d)/(e*f) which may perfectly valid results in range of a long, but the intermediate nominator or denominator does not.

You could write your own Currency class, using a long to hold the amount. The class methods would set and get the amount using a String.
Division will be a concern no matter whether you use a long or a BigDecimal. You have to determine on a case by case basis what you do with fractional cents. Discard them, round them, or save them (somewhere besides your own account).