Rounding - Knowledge

142: 4248:

correctly rounded result can be guaranteed, may demand a lot of computation time or may be out of reach. In practice, when this limit is not known (or only a very large bound is known), some decision has to be made in the implementation (see below); but according to a probabilistic model, correct rounding can be satisfied with a very high probability when using an intermediate accuracy of up to twice the number of digits of the target format plus some small constant (after taking special cases into account).

4232:, square root, and floating-point remainder will give the correctly rounded result of the infinite-precision operation. No such guarantee was given in the 1985 standard for more complex functions and they are typically only accurate to within the last bit at best. However, the 2008 standard guarantees that conforming implementations will give correctly rounded results which respect the active rounding mode; implementation of the functions, however, is optional. 1204: 1410: 43: 2718: 1023: 1247: 2211: 2017: 1810: 1625: 4506:

manufacturers. The standard allows the user to choose among several rounding modes, and in each case specifies precisely how the results should be rounded. These features made numerical computations more predictable and machine-independent, and made possible the efficient and consistent implementation of

2563: 5953:

rather than the odd number that is equally near. The reason for this procedure is that in a series of several measurements of the same quantity it will be as apt to make a record too large as it will to make one too small, and so in the average of several such values will cause but a slight error, if

3971:

Error diffusion tries to ensure the error, on average, is minimized. When dealing with a gentle slope from one to zero, the output would be zero for the first few terms until the sum of the error and the current value becomes greater than 0.5, in which case a 1 is output and the difference subtracted

2499:

One method, more obscure than most, is to alternate direction when rounding a number with 0.5 fractional part. All others are rounded to the closest integer. Whenever the fractional part is 0.5, alternate rounding up or down: for the first occurrence of a 0.5 fractional part, round up, for the second

2374:

With decimal arithmetic, final digits of 0 and 5 are avoided; if there is a choice between numbers with the least significant digit 0 or 1, 4 or 5, 5 or 6, 9 or 0, then the digit different from 0 or 5 shall be selected; otherwise, the choice is arbitrary. IBM defines that, in the latter case, a digit

724:

in 1982. It was initially set at 1000.000 (three decimal places of accuracy), and after 22 months had fallen to about 520, although the market appeared to be rising. The problem was caused by the index being recalculated thousands of times daily, and always being truncated (rounded down) to 3 decimal

5897:

even is that its sign is arbitrary, or is not fixed by the computation as is the case with all the other errors. However, the computer's rule, which makes the last rounded figure of an interpolated value even when half a unit is to be disposed of, will, in the long run, make this error as often plus

4128:

Rounding a number twice in succession to different levels of precision, with the latter precision being coarser, is not guaranteed to give the same result as rounding once to the final precision except in the case of directed rounding. For instance rounding 9.46 to one decimal gives 9.5, and then 10

2337:

This method is also free from positive/negative bias and bias toward/away from zero, provided the numbers to be rounded are neither mostly even nor mostly odd. It also shares the round half to even property of distorting the original distribution, as it increases the probability of odds relative to

4159:

In some algorithms, an intermediate result is computed in a larger precision, then must be rounded to the final precision. Double rounding can be avoided by choosing an adequate rounding for the intermediate computation. This consists in avoiding to round to midpoints for the final rounding (except

2730:

can never be. For example, suppose one started with 0 and added 0.3 to that one hundred times while rounding the running total between every addition. The result would be 0 with regular rounding, but with stochastic rounding, the expected result would be 30, which is the same value obtained without

4063:

where the rounding is randomly up or down. Stochastic rounding can be used for Monte Carlo arithmetic, but in general, just rounding up or down with equal probability is more often used. Repeated runs will give a random distribution of results which can indicate the stability of the computation.

4643:

Office of the Federal Coordinator for Meteorology determined that weather data should be rounded to the nearest round number, with the "round half up" tie-breaking rule. For example, 1.5 rounded to integer should become 2, and −1.5 should become −1. Prior to that date, the tie-breaking rule was

2237:

It is often used for currency conversions and price roundings (when the amount is first converted into the smallest significant subdivision of the currency, such as cents of a euro) as it is easy to explain by just considering the first fractional digit, independently of supplementary precision

4505:

Until the 1980s, the rounding method used in floating-point computer arithmetic was usually fixed by the hardware, poorly documented, inconsistent, and different for each brand and model of computer. This situation changed after the IEEE 754 floating-point standard was adopted by most computer

4247:

results, except on some well-known arguments; therefore, from a theoretical point of view, it is always possible to correctly round such functions. However, for an implementation of such a function, determining a limit for a given precision on how accurate results need to be computed, before a

4615:, and do not discriminate between integers and floating-point values; however, the implementations of these languages will typically convert these numbers into IEEE 754 double-precision floating-point values before exposing the computed digits with a limited precision (notably within standard 4330:'s libmcr of 2004, in the 4 rounding modes. For the difficult cases, this library also uses multiple precision, and the number of words is increased by 2 each time the Table-maker's dilemma occurs (with undefined behavior in the very unlikely event that some limit of the machine is reached). 4571:

of most languages have commonly provided at least the four basic rounding functions (up, down, to nearest, and toward zero). The tie-breaking method can vary depending on the language and version or might be selectable by the programmer. Several languages follow the lead of the IEEE 754

2266:. Thus, for example, 23.5 becomes 24, as does 24.5; however, −23.5 becomes −24, as does −24.5. This function minimizes the expected error when summing over rounded figures, even when the inputs are mostly positive or mostly negative, provided they are neither mostly even nor mostly odd. 1199:{\displaystyle y=\operatorname {truncate} (x)=\operatorname {sgn}(x)\left\lfloor \left|x\right|\right\rfloor =-\operatorname {sgn}(x)\left\lceil -\left|x\right|\right\rceil ={\begin{cases}\left\lfloor x\right\rfloor &x\geq 0\\\left\lceil x\right\rceil &x<0\end{cases}}} 1442:

If it were not for the 0.5 fractional parts, the round-off errors introduced by the round to nearest method would be symmetric: for every fraction that gets rounded down (such as 0.268), there is a complementary fraction (namely, 0.732) that gets rounded up by the same amount.

661:

The most basic form of rounding is to replace an arbitrary number by an integer. All the following rounding modes are concrete implementations of an abstract single-argument "round()" procedure. These are true functions (with the exception of those that use randomness).

4147:

standard dictate that in straightforward calculations the result should not be rounded twice. This has been a particular problem with Java as it is designed to be run identically on different machines, special programming tricks have had to be used to achieve this with

4662:

may write "−0" to indicate a temperature between 0.0 and −0.5 degrees (exclusive) that was rounded to an integer. This notation is used when the negative sign is considered important, no matter how small is the magnitude; for example, when rounding temperatures in the

4129:

when rounding to integer using rounding half to even, but would give 9 when rounded to integer directly. Borman and Chatfield discuss the implications of double rounding when comparing data rounded to one decimal place to specification limits expressed using integers.

3275:

The most common type of rounding is to round to an integer; or, more generally, to an integer multiple of some increment – such as rounding to whole tenths of seconds, hundredths of a dollar, to whole multiples of 1/2 or 1/8 inch, to whole dozens or thousands, etc.

3829:

Where the rounded result would overflow the result for a directed rounding is either the appropriate signed infinity when "rounding away from zero", or the highest representable positive finite number (or the lowest representable negative finite number if

2375:

with the smaller magnitude shall be selected. RPSP can be applied with the step between two consequent roundings as small as a single digit (for example, rounding to 1/10 can be applied after rounding to 1/100). For example, when rounding to integer,

1405:{\displaystyle y=\operatorname {sgn}(x)\left\lceil \left|x\right|\right\rceil =-\operatorname {sgn}(x)\left\lfloor -\left|x\right|\right\rfloor ={\begin{cases}\left\lceil x\right\rceil &x\geq 0\\\left\lfloor x\right\rfloor &x<0\end{cases}}} 3903:

Many design procedures describe how to calculate an approximate value, and then "round" to some standard size using phrases such as "round down to nearest standard value", "round up to nearest standard value", or "round to nearest standard value".

2082: 1888: 4140:, litigated between 1995 and 1997, the insurance companies argued that double rounding premiums was permissible and in fact required. The US courts ruled against the insurance companies and ordered them to adopt rules to ensure single rounding. 970: 3671: 2025:

This method treats positive and negative values symmetrically, and therefore is free of overall positive/negative bias if the original numbers are positive or negative with equal probability. It does, however, still have bias toward zero.

860: 2713:{\displaystyle \operatorname {Round} (x)={\begin{cases}\lfloor x\rfloor &{\text{ with probability }}1-(x-\lfloor x\rfloor )=\lfloor x\rfloor -x+1\\\lfloor x\rfloor +1&{\text{ with probability }}{x-\lfloor x\rfloor }\end{cases}}} 1690: 1505: 2234:, treats positive and negative values symmetrically, and therefore is free of overall positive/negative bias if the original numbers are positive or negative with equal probability. It does, however, still have bias away from zero. 4215:

be rounded within half an ulp like SQRT? Because nobody knows how much computation it would cost... No general way exists to predict how many extra digits will have to be carried to compute a transcendental expression and round it

641: 2507:

If occurrences of 0.5 fractional parts occur significantly more than a restart of the occurrence "counting", then it is effectively bias free. With guaranteed zero bias, it is useful if the numbers are to be summed or averaged.

4324:), first distributed in 2003. It supports the 4 rounding modes and is proved, using the knowledge of the hardest-to-round cases. More efficient than IBM MathLib. Succeeded by Metalibm (2014), which automates the formal proofs. 725:

places, in such a way that the rounding errors accumulated. Recalculating the index for the same period using rounding to the nearest thousandth rather than truncation corrected the index value from 524.811 up to 1098.892.

3410: 5948:

A fraction perceptibly less than a half should be discarded and more than a half should always be considered as one more unit, but when it is uncertain which figure is the nearer one the universally adopted rule is to

4164:. Equivalently, it consists in returning the intermediate result when it is exactly representable, and the nearest floating-point number with an odd significand otherwise; this is why it is also known as 2355:

value would round to a normal non-zero value. Effectively, this mode prefers preserving the existing scale of tie numbers, avoiding out-of-range results when possible for numeral systems of even

4160:

when the midpoint is exact). In binary arithmetic, the idea is to round the result toward zero, and set the least significant bit to 1 if the rounded result is inexact; this rounding is called

4498:

remains more obscure. If this rounding method was ever a standard in banking, the evidence has proved extremely difficult to find. To the contrary, section 2 of the European Commission report

2345:

This variant is almost never used in computations, except in situations where one wants to avoid increasing the scale of floating-point numbers, which have a limited exponent range. With

2310:

However, this rule distorts the distribution by increasing the probability of evens relative to odds. Typically this is less important than the biases that are eliminated by this method.

2223:

representation for the values to be rounded, because only the first omitted digit needs to be considered to determine if it rounds up or down. This is one method used when rounding to

2553:

Rounding as follows to one of the closest integer toward negative infinity and the closest integer toward positive infinity, with a probability dependent on the proximity is called

5030:

Bankers' rounding is used when truncating real numbers that end with .5; that is, odd numbers are rounded up to an even integer, even numbers are rounded down to an even integer.

3808:

Apart from this detail, all the variants of rounding discussed above apply to the rounding of floating-point numbers as well. The algorithm for such rounding is presented in the

2206:{\displaystyle y=\operatorname {sgn}(x)\left\lfloor \left|x\right|+{\tfrac {1}{2}}\right\rfloor =-\operatorname {sgn}(x)\left\lceil -\left|x\right|-{\tfrac {1}{2}}\right\rceil } 2012:{\displaystyle y=\operatorname {sgn}(x)\left\lceil \left|x\right|-{\tfrac {1}{2}}\right\rceil =-\operatorname {sgn}(x)\left\lfloor -\left|x\right|+{\tfrac {1}{2}}\right\rfloor } 4351:

for which a rounded value can never be determined no matter how many digits are calculated. Specific instances cannot be given but this follows from the undecidability of the

4553:, would provide only one method, usually truncation (toward zero). This default method could be implied in certain contexts, such as when assigning a fractional number to an 4564:. Other kinds of rounding had to be programmed explicitly; for example, rounding a positive number to the nearest integer could be implemented by adding 0.5 and truncating. 4313:(abbreviated as APMathLib or just MathLib), also called libultim, in rounding to nearest only. This library uses up to 768 bits of working precision. It was included in the 3872:

do not exceed a given maximum. This problem is fairly distinct from that of rounding a value to a fixed number of decimal or binary digits, or to a multiple of a given unit

1454:

fractional parts, the rounding errors by all values, with the omission of those having 0.5 fractional part, would statistically compensate each other. This means that the

3699:

on a logarithmic scale. In particular, for resistors with a 10% accuracy, they are supplied with nominal values 100, 120, 150, 180, 220, etc. rounded to multiples of 10 (

901: 3521: 3475:

The abstract single-argument "round()" function that returns an integer from an arbitrary real value has at least a dozen distinct concrete definitions presented in the

272:

is difficult because the number of extra digits that need to be calculated to resolve whether to round up or down cannot be known in advance. This problem is known as "

1805:{\displaystyle y=\left\lceil x-{\tfrac {1}{2}}\right\rceil =-\left\lfloor -x+{\tfrac {1}{2}}\right\rfloor =\left\lfloor {\tfrac {1}{2}}\lceil 2x\rceil \right\rfloor } 1620:{\displaystyle y=\left\lfloor x+{\tfrac {1}{2}}\right\rfloor =-\left\lceil -x-{\tfrac {1}{2}}\right\rceil =\left\lceil {\tfrac {1}{2}}\lfloor 2x\rfloor \right\rceil } 780: 3956:, such as sound waves, the overall effect of a number of measurements is more important than the accuracy of each individual measurement. In these circumstances, 4592:). On the other hand, truncation (round to zero) is still the default rounding method used by many languages, especially for the division of two integer values. 4584:, provide functions that round a value to a specified number of decimal digits (e.g., from 4321.5678 to 4321.57 or 4300). In addition, many languages provide a 716:

Where many calculations are done in sequence, the choice of rounding method can have a very significant effect on the result. A famous instance involved a new

4588:

or similar string formatting function, which allows one to convert a fractional number to a string, rounded to a user-specified number of decimal places (the

161: 573: 5575: 5093: 4302:

were provided. According to its documentation, this library uses a first step with an accuracy a bit larger than double precision, a second step based on

5747: 4403:

rounded to more than four digits. In contrast, truncation does not suffer from this problem; for example, a simple string search for "3.1415", which is

3715:. The logarithm of 165 is closer to the logarithm of 180 therefore a 180 ohm resistor would be the first choice if there are no other considerations. 5999: 5065:

Gupta, Suyog; Angrawl, Ankur; Gopalakrishnan, Kailash; Narayanan, Pritish (2016-02-09). "Deep Learning with Limited Numerical Precision". p. 3.

2545:

values. An advantage over alternate tie-breaking is that the last direction of rounding on the 0.5 fractional part does not have to be "remembered".

6170: 4502:

suggests that there had previously been no standard approach to rounding in banking; and it specifies that "half-way" amounts should be rounded up.

3692:. Rounding on a logarithmic scale is accomplished by taking the log of the amount and doing normal rounding to the nearest value on the log scale. 3293: 2371:. This can be achieved if all roundings except the final one are done using RPSP, and only the final rounding uses the externally requested mode. 4220:

to some preassigned number of digits. Even the fact (if true) that a finite number of extra digits will ultimately suffice may be a deep theorem.

2504:. "Up" and "down" can be any two rounding methods that oppose each other - toward and away from positive infinity or toward and away from zero. 2307:, will give a rounded result with an error that tends to grow in proportion to the square root of the number of operations rather than linearly. 5614:

de Dinechin, Florent; Lauter, Christoph; Muller, Jean-Michel (January–March 2007). "Fast and correctly rounded logarithms in double-precision".

3486:

for the increment and then reduces to the equivalent abstract single-argument function, with also the same dozen distinct concrete definitions.

4744: 4713: 4573: 4256: 3479:

section. The abstract two-argument "roundToMultiple()" function is formally defined here, but in many cases it is used with the implicit value

1458:

of the rounded numbers is equal to the expected value of the original numbers when numbers with fractional part 0.5 from the set are removed.

566:

Approximating each of a finite set of real numbers by an integer so that the sum of the rounded numbers equals the rounded sum of the numbers

5478: 4880: 4805: 4576:

argument and returning the result of the same type, which then may be converted to an integer if necessary. This approach may avoid spurious

4072:

It is possible to use rounded arithmetic to evaluate the exact value of a function with integer domain and range. For example, if an integer

202:

Rounding is often done to obtain a value that is easier to report and communicate than the original. Rounding can also be important to avoid

3507:; for example, it is common in computing to need to round a number to a whole power of 2. The steps, in general, to round a positive number 4984: 4387:

exists and 1 otherwise. The value before rounding can however be approximated to any given precision even if the conjecture is unprovable.

376:

A rounding method should have utility in computer science or human arithmetic where finite precision is used, and speed is a consideration.

4255:

package gives correctly rounded arbitrary precision results. Some other libraries implement elementary functions with correct rounding in

387:; i.e., once a number has been rounded, rounding it again to the same precision will not change its value. Rounding functions are also 4861: 3935:, which use notions of distance other than the simple difference – for example, a sequence may round to the integer with the smallest 288: 4240: 2541:

Like round-half-to-even and round-half-to-odd, this rule is essentially free of overall bias, but it is also fair among even and odd

5210: 5010: 126: 5523: 5449:

Muller, Jean-Michel; Brisebarre, Nicolas; de Dinechin, Florent; Jeannerod, Claude-Pierre; Lefèvre, Vincent; Melquiond, Guillaume;

5132: 225:

in the reported result. Rounding is almost unavoidable when reporting many computations – especially when dividing two numbers in

4768:

A case where double rounding always leads to the same value as directly rounding to the final precision is when the radix is odd.

280: 2461:

For correct results, each rounding step must remove at least 2 binary digits, otherwise, wrong results may appear. For example,

4278:

and Moshe Olshansky in 1999, correctly rounded to nearest only. This library was claimed to be portable, but only binaries for

5593: 6175: 5972:

Selected Techniques of Statistical Analysis for Scientific and Industrial Research, and Production and Management Engineering

5834: 5002: 4608: 4449: 380:

Because it is not usually possible for a method to satisfy all ideal characteristics, many different rounding methods exist.

64: 4236: 4152:

floating point. The Java language was changed to allow different results where the difference does not matter and require a

3973: 2735:

where the training may use low precision arithmetic iteratively. Stochastic rounding is also a way to achieve 1-dimensional

701:

is negative, round-down is the same as round-away-from-zero, and round-up is the same as round-toward-zero. In any case, if

4156:

qualifier to be used when the results have to conform accurately; strict floating point has been restored in Java 17.

369:

that already exist between the domain and range. With finite precision (or a discrete domain), this translates to removing

1451: 4545:

provide functions or special syntax to round fractional numbers in various ways. The earliest numeric languages, such as

4399:

rounded to four digits is "3.1416" but a simple search for this string will not discover "3.14159" or any other value of

6165: 4752: 4201:

two floating-point arguments at which it does not over/underflow. Instead, reputable math libraries compute elementary

3968:

is used to achieve analog type output from an inertial device by rapidly pulsing the power with a variable duty cycle.

107: 4719: 879: 758: 396: 331:

As a result of (1) and (2), the output from rounding should be close to its input, often as close as possible by some

79: 5659: 5017:

Rounding to the nearest even number is also called 'bankers rounding' because the banks use this technique as well.

4554: 258: 6141:

that is accessible to a general audience but especially useful to those studying computer science and electronics.

5541: 5992: 4708: 3770: 1462: 683: 295: 250: 60: 31: 6099: 5308:"A mechanically checked proof of the correctness of the kernel of the AMD5K86 floating-point division algorithm" 4603:

do not define any specific maximum precision for numbers and measurements, which they treat and expose in their

4306:, and a third step with a 768-bit precision based on arrays of IEEE 754 double-precision floating-point numbers. 2500:

occurrence, round down, and so on. Alternatively, the first 0.5 fractional part rounding can be determined by a

86: 5918:

Here we have a case in which the half of an odd number is required. A good rule to adopt in such a case is to

5842: 5579: 4748: 4600: 4550: 4303: 721: 697:

is positive, round-down is the same as round-toward-zero, and round-up is the same as round-away-from-zero. If

392: 4688: 4356: 2475:

If the erroneous middle step is removed, the final rounding to integer rounds 3.25 to the correct value of 3.

5751: 2590: 1344: 1138: 6075: 4468: 4202: 4184: 1465:

numbers are typically used, which have even more computational nuances because they are not equally spaced.

366: 311: 269: 53: 5086: 4229: 141: 5623: 5322: 5025: 4596: 4557: 4428: 4206: 4144: 3965: 2727: 1447: 322: 230: 211: 93: 3881: 5913:

Logarithmic and Other Mathematical Tables with Examples of their Use and Hints on the Art of Computation

5864:

Standard Practice for Using Significant Digits in Test Data to Determine Conformance with Specifications

4954: 4604: 4244: 234: 5839:

The standard arose from a committee of the ASA working to standardize inch–millimeter conversion. See:

5050: 3464:

of 10, like 1/1000 or 25/100). For intermediate values stored in digital computers, it often means the

310:) is sometimes used to indicate rounding of exact numbers, e.g. 9.98 ≈ 10. This sign was introduced by 206:

reporting of a computed number, measurement, or estimate; for example, a quantity that was computed as

5307: 2238:

digits or sign of the amount (for strict equivalence between the paying and recipient of the amount).

5425: 4542: 4419:

The concept of rounding is very old, perhaps older than the concept of division itself. Some ancient

349: 332: 75: 5628: 3900:, writing paper, capacitors, and many other products are usually sold in only a few standard sizes. 965:{\displaystyle y=\operatorname {ceil} (x)=\left\lceil x\right\rceil =-\left\lfloor -x\right\rfloor } 6160: 5327: 4698: 4568: 4507: 4488: 4333:

The CORE-MATH project (2022) provides some correctly rounded functions in the 4 rounding modes for

4168:. A concrete implementation of this approach, for binary and decimal arithmetic, is implemented as 3666:{\displaystyle \mathrm {roundToPower} (x,b)=b^{\mathrm {round} (\log _{b}x)},x>0,b>0,b\neq 1} 3461: 3434: 2224: 1647: 687: 532: 478: 341: 254: 145: 6138: 5911: 5886: 5268: 6057: 5941: 5641: 5383: 5249: 5066: 4905: 4838: 4789: 4060: 3885: 2876: 2866: 2856: 2322:, a similar tie-breaking rule to round half to even. In this approach, if the fractional part of 388: 5853: 4971: 4522: 3676:

Many of the caveats applicable to rounding to a multiple are applicable to rounding to a power.

5818: 3927:

More general rounding rules can separate values at arbitrary break points, used for example in

855:{\displaystyle y=\mathrm {floor} (x)=\left\lfloor x\right\rfloor =-\left\lceil -x\right\rceil } 6121: 5484: 5474: 5206: 5202: 5175: 5128: 5006: 4876: 4818: 4801: 4348: 3953: 3932: 2405:

With binary arithmetic, this rounding is also called "round to odd" (not to be confused with "

2402:" section, rounding 9.46 to one decimal gives 9.4, which rounding to integer in turn gives 9. 370: 284: 4580:

because floating-point types have a larger range than integer types. Some languages, such as

4035:. The first of these and the differences of adjacent values give the desired rounded values: 183:

value that has a shorter, simpler, or more explicit representation. For example, replacing $

6047: 6039: 5867: 5729: 5633: 5466: 5375: 5332: 5239: 5167: 4897: 4868: 4860: 4830: 4577: 4521:

in many studies, to evaluate the numeracy level of ancient populations. He came up with the

4327: 4180: 3696: 2732: 2334:. Thus, for example, 23.5 becomes 23, as does 22.5; while −23.5 becomes −23, as does −22.5. 1481:), a tie-breaking rule that is widely used in many disciplines. That is, half-way values of 5269:"Efficiently producing default orthogonal IEEE double results using extended IEEE hardware" 3834:

is negative), when "rounding toward zero". The result of an overflow for the usual case of

2723:

For example, 1.6 would be rounded to 1 with probability 0.4 and to 2 with probability 0.6.

2042:), a tie-breaking rule that is commonly taught and used, namely: If the fractional part of 325:. This way, when the same input is rounded in different instances, the output is unchanged. 6103: 6024: 5127:

Bruce Trump, Christine Schneider. "Excel Formula Calculates Standard 1%-Resistor Values".

4856: 4756: 4561: 4352: 4317:

in 2001, but the "slow paths" (providing correct rounding) were removed from 2018 to 2021.

3961: 3912: 3908: 2871: 2861: 2851: 446: 262: 222: 203: 5407: 5429: 5361:"Emulation of a FMA and correctly-rounded sums: proved algorithms using rounding to odd" 5158:

Borman, Phil; Chatfield, Marion (2015-11-10). "Avoid the perils of using rounded data".

17: 5527: 5450: 4793: 4472: 4291: 3877: 3449: 2339: 2220: 1455: 486: 464: 5360: 2303:

By eliminating bias, repeated addition or subtraction of independent numbers, as in a

636:{\displaystyle {\bigl \{}{\tfrac {3}{12}},{\tfrac {4}{12}},{\tfrac {5}{12}}{\bigr \}}} 100: 6154: 5967: 5462: 5195: 5145:"Monte Carlo Arithmetic: a framework for the statistical analysis of roundoff errors" 5144: 4703: 4676: 4659: 4640: 4514: 4491:

indicated the practice was already "well established" in data analysis by the 1940s.

4476: 4314: 3465: 737: 180: 6061: 5253: 4842: 391:; i.e., rounding two numbers to the same absolute precision will not exchange their 5645: 5502: 5387: 5356: 5116: 4901: 4682: 4631:

Some disciplines or institutions have issued standards or directives for rounding.

3928: 353: 5891:. Mathematical Monographs. Vol. 7. New York: J. Wiley & Son. p. 42. 5733: 2254:

without bias toward/away from zero. By this convention, if the fractional part of

156:

using different methods. For clarity, the graphs are shown displaced from integer

6025:"Quantifying Quantitative Literacy: Age Heaping and the History of Human Capital" 3441:

depends on the magnitude of the number to be rounded (or of the rounded result).

1818:

Some programming languages (such as Java and Python) use "half down" to refer to

395:(but may give the same value). In the general case of a discrete range, they are 4872: 4693: 4653: 4518: 4471:

called this "the computer's rule", indicating that it was then in common use by

4438: 4424: 4420: 4275: 3452:

is used to represent the numbers. For display to humans, that usually means the

2501: 2304: 1646:

This method only requires checking one digit to determine rounding direction in

717: 384: 328:

Calculations done with rounding should be close to those done without rounding.

238: 42: 5171: 4379:

is the first even number greater than 4 which is not the sum of two primes, or

4076:

is known to be a perfect square, its square root can be computed by converting

1633:

Some programming languages (such as Java and Python) use "half up" to refer to

6124: 6043: 5470: 4724: 4716:, an application of rounding to integers that has been thoroughly investigated 4620: 4616: 4360: 4299: 4117: 3700: 2554: 1435:

is exactly half-way between two integers – that is, when the fraction part of

989: 4930: 4916: 4513:

Currently, much research tends to round to multiples of 5 or 2. For example,

4437:, the length of the year, and the length of the month are also ancient – see 4183:

coined the term "The Table-Maker's Dilemma" for the unknown cost of rounding

3405:{\displaystyle \mathrm {roundToMultiple} (x,m)=\mathrm {round} (x/m)\times m} 2367:

This rounding mode is used to avoid getting a potentially wrong result after

6129: 4834: 4487:

called it a "universally adopted rule" for recording physical measurements.

3957: 2736: 1431:

to the nearest integer requires some tie-breaking rule for those cases when

242: 5696: 5179: 4529:

among regions possible without any historical sources where the population

5781: 5663: 5637: 5379: 5244: 5227: 6144: 4612: 4530: 4526: 4395:

Rounding can adversely affect a string search for a number. For example,

4321: 4252: 4228:

floating-point standard guarantees that add, subtract, multiply, divide,

4225: 4153: 2538:, with equal probability. All others are rounded to the closest integer. 2351: 2297: 192: 5702: 5558: 4383:=1 if there is no such number. The rounded result is 2 if such a number 5764: 5728:. Mathematical Software – ICMS 2014. Vol. 8592. pp. 713–717. 5545: 4909: 4664: 4546: 4475:

who calculated mathematical tables. For example, it was recommended in

4343:

libc provides some correctly rounded functions in the 4 rounding modes.

4283: 4279: 3453: 3422:

dollars to whole cents (i.e., to a multiple of 0.01) entails computing

2483: 2479: 528: 506: 358: 226: 5336: 5022:

Microsoft Pascal Compiler for the MS-DOS Operating System User's Guide

3703:). If a calculation indicates a resistor of 165 ohms is required then 214:

only to within a few hundred units is usually better stated as "about

6096: 6052: 5871: 5768: 5680: 5676: 5562: 5454: 4888:

Nievergelt, Yves (2000). "Rounding Errors to Knock Your Stocks Off".

4585: 4337:

processors. Proved using the knowledge of the hardest-to-round cases.

4334: 3897: 2216:

For example, 23.5 gets rounded to 24, and −23.5 gets rounded to −24.

2022:

For example, 23.5 gets rounded to 23, and −23.5 gets rounded to −23.

1815:

For example, 23.5 gets rounded to 23, and −23.5 gets rounded to −24.

1630:

For example, 23.5 gets rounded to 24, and −23.5 gets rounded to −23.

1415:

For example, 23.2 gets rounded to 24, and −23.2 gets rounded to −24.

1209:

For example, 23.7 gets rounded to 23, and −23.7 gets rounded to −23.

975:

For example, 23.2 gets rounded to 24, and −23.7 gets rounded to −23.

865:

For example, 23.7 gets rounded to 23, and −23.2 gets rounded to −24.

345: 176: 160:

values. In the SVG file, hover over a method to highlight it and, in

5723: 4243:, many of the standard elementary functions can be proved to return 301: 5993:"The Introduction of the Euro and the Rounding of Currency Amounts" 5576:"libultim – ultimate correctly-rounded elementary-function library" 5286: 5071: 2382:

20.01, 20.1, 20.9, 20.99, 21, 21.01, 21.9, 21.99 are rounded to 21;

5488: 4287: 3802: 2356: 4917:"Ever had problems rounding off figures? This stock exchange has" 4500:

The Introduction of the Euro and the Rounding of Currency Amounts

682:

are all directed toward or away from the same limiting value (0,

257:. In a sequence of calculations, these rounding errors generally 221:

On the other hand, rounding of exact numbers will introduce some

4572:

floating-point standard, and define these functions as taking a

4453: 4340: 246: 4821:(July 1977). "Mathematical foundation of computer arithmetic". 3976:

is a popular error diffusion procedure when digitizing images.

3911:

is equally spaced on a logarithmic scale, choosing the closest

3753:. The value 165 rounds to 180 in the resistors example because 5695:

Sibidanov, Alexei; Zimmermann, Paul; Glondu, Stéphane (2022).

5143:

Parker, D. Stott; Eggert, Paul R.; Pierce, Brad (2000-03-28).

4581: 4295: 4263: 4169: 4149: 36: 5946:. Philadelphia: Jefferson Laboratory of Physics. p. 29. 5801: 5197:

Class Action Dilemmas: Pursuing Public Goals for Private Gain

4972:"decimal – Decimal fixed point and floating point arithmetic" 2726:

Stochastic rounding can be accurate in a way that a rounding

2349:, a non-infinite number would round to infinity, and a small 5503:"NA Digest Sunday, April 18, 1999 Volume 99 : Issue 16" 5847:

Industrial Standardization and Commercial Standards Monthly

4999:

Postcards 4 Language Booster: Workbook with Grammar Builder

4464:

are fairly self-explanatory. In the 1906 fourth edition of

4432: 2706: 2269:

This variant of the round-to-nearest method is also called

1398: 1192: 430: 4798:

Linear Algebra as an Introduction to Abstract Mathematics.

4567:

In the last decades, however, the syntax and the standard

3995:

occur in order and each is to be rounded to a multiple of

552:

Approximating a value by a multiple of a specified amount

5287:"JEP 306: Restore Always-Strict Floating-Point Semantics" 4654:

Signed zero § In rounded values such as temperatures

3846:

In some contexts it is desirable to round a given number

2300:

operations for results in binary floating-point formats.

1229:

is the integer that is closest to 0 (or equivalently, to

5748:"libmcr – correctly-rounded elementary-function library" 3805:(usually 2 or 10) of the floating-point representation. 2409:"). For example, when rounding to 1/4 (0.01 in binary), 5703:

29th IEEE Symposium on Computer Arithmetic (ARITH 2022)

5306:

Moore, J. Strother; Lynch, Tom; Kaufmann, Matt (1996).

4679:, dealing with the absence of extremely low-value coins 4431:

and square roots in base 60. Rounded approximations to

2338:

evens. It was the method used for bank balances in the

650:

Sum of rounded elements equals rounded sum of elements

4997:

Abbs, Brian; Barker, Chris; Freebairn, Ingrid (2003).

4251:

Some programming packages offer correct rounding. The

4104:

is not too big, the floating-point round-off error in

2557:

rounding and will give an unbiased result on average.

2187: 2127: 1993: 1933: 1774: 1749: 1712: 1589: 1564: 1527: 615: 600: 585: 3850:

to a "neat" fraction – that is, the nearest fraction

3524: 3296: 2566: 2250:, a tie-breaking rule without positive/negative bias 2085: 1891: 1693: 1508: 1250: 1026: 904: 783: 576: 5851:

The standard was also more concisely advertised in:

5893:An important fact with regard to the error 1/2 for 4363:, then the result of rounding the following value, 3426:, then rounding that to 218, and finally computing 2743:

Comparison of approaches for rounding to an integer

317:Ideal characteristics of rounding methods include: 67:. Unsourced material may be challenged and removed. 5860:. Vol. 84, no. 11. Nov 1940. p. 93. 5690: 5688: 5194: 4456:standard E-29 since 1940. The origin of the terms 4407:truncated to four digits, will discover values of 3979:As a one-dimensional example, suppose the numbers 3919:. Such rounded values can be directly calculated. 3812:section above, but with a constant scaling factor 3689: 3665: 3404: 2712: 2205: 2011: 1804: 1619: 1404: 1198: 964: 854: 635: 164:-enabled browsers, click to select or deselect it. 5725:Metalibm: A Mathematical Functions Code Generator 5660:"CRlibm – Correctly Rounded mathematical library" 5160:Journal of Pharmaceutical and Biomedical Analysis 4560:, or using a fractional number as an index of an 4209:and almost always well within one ulp. Why can't 2471:3.5 round-half-to-even to 1 ⇒ result is 4 (wrong) 2219:This can be more efficient on computers that use 1635: 1485:are always rounded up. If the fractional part of 690:and is often required in financial calculations. 427:Approximating an irrational number by a fraction 6139:An introduction to different rounding algorithms 5819:Duncan J. Melville. "YBC 7289 clay tablet". 2006 4931:"Vancouver stock index has right number at last" 3964:, are normally used. A related technique called 674:, as the displacements from the original number 5455:"Chapter 12: Solving the Table Maker's Dilemma" 4320:CRlibm, written in the old Arénaire team (LIP, 4191:Nobody knows how much it would cost to compute 4189: 3503:is very different from rounding to a specified 3283:to a multiple of some specified positive value 5147:. IEEE Computation in Science and Engineering. 4863:Accuracy and stability of numerical algorithms 3931:. A related mathematically formalized tool is 1820: 891:is the smallest integer that is not less than 273: 5849:. Vol. 11, no. 9. pp. 230–233. 5722:Kupriianova, Olga; Lauter, Christoph (2014). 5060: 5058: 628: 579: 8: 6076:"ECMA-262 ECMAScript Language Specification" 4667:scale, where below zero indicates freezing. 4108:will be less than 0.5, so the rounded value 3915:to any given value can be seen as a form of 2699: 2693: 2670: 2664: 2645: 2639: 2630: 2624: 2599: 2593: 1794: 1785: 1609: 1600: 770:is the largest integer that does not exceed 265:cases they may make the result meaningless. 195:312/937 with 1/3, or the expression √2 with 6145:How To Implement Custom Rounding Procedures 5041:Schedule 1 of the Decimal Currency Act 1969 4537:Rounding functions in programming languages 3684:This type of rounding, which is also named 3433:When rounding to a predetermined number of 5051:IBM z/Architecture Principles of Operation 4013:2.9423 = 0.9677 + 0.9204 + 0.7451 + 0.3091 3448:is normally a finite fraction in whatever 2388:24.0, 24.1, 24.9, 24.99 are rounded to 24; 2385:22.0, 22.1, 22.9, 22.99 are rounded to 22; 2296:This is the default rounding mode used in 6051: 5974:. New York: McGraw-Hill. pp. 187–223 5627: 5326: 5243: 5070: 4481:Logarithmic and Other Mathematical Tables 4205:mostly within slightly more than half an 4170:Rounding to prepare for shorter precision 4068:Exact computation with rounded arithmetic 4059:Monte Carlo arithmetic is a technique in 3747:is greater than or less than the product 3695:For example, resistors are supplied with 3610: 3586: 3585: 3525: 3523: 3385: 3362: 3297: 3295: 2686: 2681: 2604: 2585: 2565: 2406: 2363:Rounding to prepare for shorter precision 2186: 2126: 2084: 1992: 1932: 1890: 1773: 1748: 1711: 1692: 1588: 1563: 1526: 1507: 1339: 1249: 1133: 1025: 903: 790: 782: 627: 626: 614: 599: 584: 578: 577: 575: 463:Approximating a fraction by a fractional 127:Learn how and when to remove this message 5916:. New York: Henry Holt. pp. 14–15. 5542:"Accurate Portable Mathematical Library" 5453:; Stehlé, Damien; Torres, Serge (2010). 4966: 4964: 4084:, computing the approximate square root 3476: 2746: 409: 140: 5970:. In Eisenhart; Hastay; Wallis (eds.). 5831:Rules for Rounding Off Numerical Values 4781: 4736: 4639:In a guideline issued in mid-1966, the 4092:with floating point, and then rounding 3741:depends upon whether the squared value 2767:Round to prepare for shorter precision 2478:RPSP is implemented in hardware in IBM 686:, or −∞). Directed rounding is used in 509:by an integer with more trailing zeros 449:by a fraction with smaller denominator 27:Replacing a number with a simpler value 5968:"Effects of Rounding or Grouping Data" 4949: 4947: 4714:Party-list proportional representation 4525:, which enables the comparison of the 4427:contain tables with rounded values of 4311:Accurate portable mathematical library 3773:, rounding aims to turn a given value 279:Rounding has many similarities to the 253:representation with a fixed number of 6097:Federal Meteorological Handbook No. 1 5802:"Math Functions — The LLVM C Library" 5459:Handbook of Floating-Point Arithmetic 5408:"21718 – real.c rounding not perfect" 4985:Engineering Drafting Standards Manual 356:. A classical range is the integers, 270:transcendental mathematical functions 7: 5228:"When is double rounding innocuous?" 4611:interface as strings as if they had 4411:truncated to more than four digits. 4120:could be used for exact arithmetic. 4015:, are each rounded to a multiple of 3999:. In this case the cumulative sums, 3916: 3838:is always the appropriate infinity. 3809: 2399: 2368: 65:adding citations to reliable sources 3511:to a power of some positive number 3472:is an integer times a power of 2). 1826:round half toward negative infinity 1664:round half toward negative infinity 1641:round half toward positive infinity 1479:round half toward positive infinity 407:Typical rounding problems include: 4755:, and for distributing the total 3599: 3596: 3593: 3590: 3587: 3559: 3556: 3553: 3550: 3547: 3544: 3541: 3538: 3535: 3532: 3529: 3526: 3375: 3372: 3369: 3366: 3363: 3340: 3337: 3334: 3331: 3328: 3325: 3322: 3319: 3316: 3313: 3310: 3307: 3304: 3301: 3298: 2465:3.125 RPSP to 1/4 ⇒ result is 3.25 2448:⇒ result is 2.75 (10.11 in binary) 2429:⇒ result is 2.25 (10.01 in binary) 2342:when it decimalized its currency. 1001:is the integer that is closest to 803: 800: 797: 794: 791: 25: 4800:World Scientific, Singapur 2016, 4112:will be the exact square root of 4009:2.6332 = 0.9677 + 0.9204 + 0.7451 3793:that depends on the magnitude of 3789:should be a multiple of a number 2490:Randomized rounding to an integer 2438:⇒ result is 2.5 (10.10 in binary) 1844:) as opposed to the conventional 5888:Probability and theory of errors 5430:"A Logarithm Too Clever by Half" 5226:Samuel A. Figueroa (July 1995). 5087:"Zener Diode Voltage Regulators" 4466:Probability and Theory of Errors 4391:Interaction with string searches 4368: 4143:Some computer languages and the 3271:Rounding to a specified multiple 2731:rounding. This can be useful in 2468:3.25 RPSP to 1/2 ⇒ result is 3.5 1666:) as opposed to the more common 740:applied to the original number, 41: 6171:Statistical data transformation 6005:from the original on 2010-10-09 5359:; Melquiond, Guillaume (2008). 5267:Roger Golliver (October 1998). 5099:from the original on 2011-07-13 5024:. Microsoft Corporation. 1985. 3686:rounding to a logarithmic scale 2457:⇒ result is 3 (11.00 in binary) 2419:⇒ result is 2 (10.00 in binary) 2359:(such as binary and decimal).. 2262:is the even integer nearest to 1425:rounding to the nearest integer 1419:Rounding to the nearest integer 1017:, without its fraction digits. 672:directed rounding to an integer 666:Directed rounding to an integer 383:As a general rule, rounding is 340:To be considered rounding, the 52:needs additional citations for 5951:record the nearest even number 5835:American Standards Association 5368:IEEE Transactions on Computers 5315:IEEE Transactions on Computers 4902:10.1080/0025570X.2000.11996800 4823:IEEE Transactions on Computers 3892:Rounding to an available value 3622: 3603: 3575: 3563: 3393: 3379: 3356: 3344: 3279:In general, rounding a number 2633: 2615: 2579: 2573: 2394:25.01, 25.1 are rounded to 26. 2330:is the odd integer nearest to 2161: 2155: 2104: 2098: 1967: 1961: 1910: 1904: 1309: 1303: 1269: 1263: 1103: 1097: 1063: 1057: 1045: 1039: 1005:such that it is between 0 and 923: 917: 885:round toward positive infinity 813: 807: 764:round toward negative infinity 670:These four methods are called 527:Approximating a large decimal 287:must be encoded by numbers or 1: 5843:"Man's Love Of Round Numbers" 5734:10.1007/978-3-662-44199-2_106 4974:. Python Software Foundation. 4743:This is needed e.g. for the 4644:"round half away from zero". 4371:cannot be determined: either 4272:Mathematical Library for Java 4241:Lindemann–Weierstrass theorem 3948:Dithering and error diffusion 3876:. This problem is related to 3842:Rounding to a simple fraction 3690:rounding to a specified power 3495:Rounding to a specified power 3287:entails the following steps: 1842:round half away from infinity 1650:and similar representations. 1446:When rounding a large set of 1423:These six methods are called 321:Rounding should be done by a 5966:Churchill Eisenhart (1947). 5885:Woodward, Robert S. (1906). 4929:Lilley, Wayne (1983-11-29). 4867:(2nd ed.). p. 54. 4753:Mathematics of apportionment 4648:Negative zero in meteorology 2683: with probability 2606: with probability 2030:Rounding half away from zero 1848:. If the fractional part of 1670:. If the fractional part of 397:piecewise constant functions 6032:Journal of Economic History 5193:Deborah R. Hensler (2000). 4915:Quinn, Kevin (1983-11-08). 4873:10.1137/1.9780898718027.ch2 4759:of an invoice to its items) 4720:Signed-digit representation 3960:, and a related technique, 3781:with a specified number of 2305:one-dimensional random walk 2230:This method, also known as 485:Approximating a fractional 6192: 6106:, Washington, DC., 104 pp. 5943:The Theory of Measurements 5172:10.1016/j.jpba.2015.07.021 4747:, implemented e.g. by the 4651: 4116:. This is essentially why 4080:to a floating-point value 3943:Rounding in other contexts 2752: 2516:If the fractional part of 2040:round half toward infinity 29: 6044:10.1017/S0022050709001120 5841:Agnew, P. G. (Sep 1940). 5471:10.1007/978-0-8176-4705-6 4709:Kahan summation algorithm 4257:IEEE 754 double precision 4237:Gelfond–Schneider theorem 3974:Floyd–Steinberg dithering 3771:floating-point arithmetic 3208: 3205: 3202: 3188: 3185: 3182: 3179: 3165: 3162: 3159: 3156: 3153: 3150: 3147: 3121: 3118: 3115: 3101: 3098: 3095: 3092: 3078: 3075: 3072: 3069: 3066: 3063: 3060: 3034: 3031: 3028: 3014: 3011: 3008: 3005: 2991: 2988: 2985: 2982: 2979: 2976: 2973: 2947: 2944: 2941: 2915: 2909: 2906: 2900: 2894: 2891: 2888: 2885: 2775: 2772: 2769: 2766: 2763: 2760: 2755: 2749: 2036:round half away from zero 1846:round half away from zero 1832:Rounding half toward zero 1636:round half away from zero 489:by one with fewer digits 365:Rounding should preserve 274:the table-maker's dilemma 32:Rounding (disambiguation) 5594:"Git - glibc.git/commit" 4987:(NASA), X-673-64-1F, p90 4955:"java.math.RoundingMode" 4749:largest remainder method 4627:Other rounding standards 4304:double-double arithmetic 4203:transcendental functions 4185:transcendental functions 4005:1.8881 = 0.9677 + 0.9204 3785:digits. In other words, 3499:Rounding to a specified 3266:Rounding to other values 2495:Alternating tie-breaking 1456:expected (average) value 995:round away from infinity 728:For the examples below, 722:Vancouver Stock Exchange 505:Approximating a decimal 18:Nearest integer function 5940:Tuttle, Lucius (1916). 5910:Newcomb, Simon (1882). 5854:"Rounding Off Decimals" 5782:"The CORE-MATH project" 5524:"Math Library for Java" 5117:"Build a Mirror Tester" 4835:10.1109/TC.1977.1674893 4635:US weather observations 4494:The origin of the term 4483:. Lucius Tuttle's 1916 4469:Robert Simpson Woodward 4462:statistician's rounding 4175: 4096:to the nearest integer 3972:from the error so far. 3765:Floating-point rounding 2275:statistician's rounding 2227:due to its simplicity. 1213:Rounding away from zero 1013:is the integer part of 312:Alfred George Greenhill 6083:ecma-international.org 5616:RAIRO-Theor. Inf. Appl 4745:apportionment of seats 4574:double-precision float 4485:Theory of Measurements 4369:up to the next integer 4222: 4197:correctly rounded for 4055:Monte Carlo arithmetic 3966:pulse-width modulation 3819:, and an integer base 3667: 3460:is an integer times a 3454:decimal numeral system 3424:2.1784 / 0.01 = 217.84 3415:For example, rounding 3406: 2770:Alternating tie-break 2714: 2391:25.0 is rounded to 25; 2379:20.0 is rounded to 20; 2207: 2013: 1838:round half toward zero 1821:round half toward zero 1806: 1621: 1406: 1200: 966: 856: 637: 547:3 significant figures 522:3 significant figures 308:approximately equal to 235:mathematical functions 231:fixed-point arithmetic 165: 6176:Theory of computation 6147:by Microsoft (broken) 5698:The CORE-MATH Project 5426:Kahan, William Morton 5380:10.1109/TC.2007.70819 5245:10.1145/221332.221334 5232:ACM SIGNUM Newsletter 5001:. Pearson Education. 4857:Higham, Nicholas John 4689:Gal's accurate tables 4685:, a similar operation 4623:interface bindings). 4543:programming languages 4448:method has served as 4357:Goldbach's conjecture 4176:Table-maker's dilemma 3668: 3466:binary numeral system 3407: 2715: 2398:In the example from " 2242:Rounding half to even 2208: 2046:is exactly 0.5, then 2014: 1852:is exactly 0.5, then 1807: 1674:is exactly 0.5, then 1622: 1489:is exactly 0.5, then 1452:uniformly distributed 1407: 1223:round toward infinity 1201: 967: 857: 678:to the rounded value 638: 268:Accurate rounding of 144: 6023:Baten, Jörg (2009). 4890:Mathematics Magazine 4134:Martinez v. Allstate 3522: 3490:Logarithmic rounding 3294: 2564: 2314:Rounding half to odd 2083: 1889: 1691: 1506: 1427:. Rounding a number 1248: 1219:round away from zero 1024: 979:Rounding toward zero 902: 781: 574: 561:Multiple of 15 458:1-digit-denominator 440:1-digit-denominator 204:misleadingly precise 61:improve this article 30:For other uses, see 6166:Computer arithmetic 5837:. 1940. Z25.1-1940. 5638:10.1051/ita:2007003 4924:Wall Street Journal 4699:Interval arithmetic 4508:interval arithmetic 4489:Churchill Eisenhart 4355:. For instance, if 4270:, which stands for 4061:Monte Carlo methods 3886:continued fractions 3515:other than 1, are: 3477:rounding to integer 2756:Randomized methods 2753:Functional methods 2549:Stochastic rounding 2512:Random tie-breaking 2271:convergent rounding 2232:commercial rounding 2225:significant figures 688:interval arithmetic 657:Rounding to integer 533:scientific notation 422:Rounding criterion 285:physical quantities 210:but is known to be 6122:Weisstein, Eric W. 6102:1999-04-20 at the 4819:Kulisch, Ulrich W. 4790:Bruno Nachtergaele 4613:infinite precision 4446:round-half-to-even 4349:computable numbers 4230:fused multiply–add 4138:Sendejo v. Farmers 4049:0.31 = 2.94 − 2.63 4045:0.74 = 2.63 − 1.89 4041:0.92 = 1.89 − 0.97 3954:continuous signals 3933:signpost sequences 3801:is a power of the 3688:, is a variant of 3663: 3435:significant digits 3402: 2761:Directed rounding 2710: 2705: 2369:multiple roundings 2347:round half to even 2248:round half to even 2203: 2196: 2136: 2009: 2002: 1942: 1802: 1783: 1758: 1721: 1654:Rounding half down 1617: 1598: 1573: 1536: 1402: 1397: 1196: 1191: 962: 852: 633: 624: 609: 594: 255:significant digits 249:; or when using a 175:means replacing a 166: 5920:write the nearest 5480:978-0-8176-4704-9 5337:10.1109/12.713311 5285:Darcy, Joseph D. 5238:(3). ACM: 21–25. 5201:. RAND. pp. 5129:Electronic Design 4882:978-0-89871-521-7 4806:978-981-4730-35-8 4496:bankers' rounding 4458:unbiased rounding 4450:American Standard 3939:(percent) error. 3882:Stern–Brocot tree 3759:150 × 180 = 27000 3697:preferred numbers 3428:218 × 0.01 = 2.18 3263: 3262: 2842:Half Away From 0 2836: 2823: 2804: 2791: 2773:Random tie-break 2764:Round to nearest 2684: 2607: 2524:randomly between 2407:round half to odd 2320:round half to odd 2291:bankers' rounding 2287:odd–even rounding 2283:Gaussian rounding 2195: 2135: 2061:is positive, and 2001: 1941: 1867:is positive, and 1782: 1757: 1720: 1597: 1572: 1535: 1237:is between 0 and 1009:(included); i.e. 985:round toward zero 654: 653: 623: 608: 593: 500:2 decimal places 413:Rounding problem 403:Types of rounding 283:that occurs when 261:, and in certain 233:; when computing 137: 136: 129: 111: 16:(Redirected from 6183: 6135: 6134: 6107: 6093: 6087: 6086: 6080: 6072: 6066: 6065: 6055: 6029: 6020: 6014: 6013: 6011: 6010: 6004: 5997: 5989: 5983: 5982: 5980: 5979: 5963: 5957: 5956: 5937: 5931: 5930: 5907: 5901: 5900: 5896: 5882: 5876: 5875: 5872:10.1520/E0029-13 5861: 5850: 5838: 5827: 5821: 5816: 5810: 5809: 5798: 5792: 5791: 5789: 5788: 5778: 5772: 5762: 5756: 5755: 5750:. Archived from 5744: 5738: 5737: 5719: 5713: 5712: 5710: 5709: 5692: 5683: 5674: 5668: 5667: 5662:. Archived from 5656: 5650: 5649: 5631: 5611: 5605: 5604: 5602: 5601: 5596:. Sourceware.org 5590: 5584: 5583: 5578:. Archived from 5572: 5566: 5556: 5550: 5549: 5544:. Archived from 5538: 5532: 5531: 5526:. Archived from 5520: 5514: 5513: 5511: 5510: 5499: 5493: 5492: 5446: 5440: 5439: 5437: 5436: 5422: 5416: 5415: 5404: 5398: 5397: 5395: 5394: 5365: 5353: 5347: 5346: 5344: 5343: 5330: 5312: 5303: 5297: 5296: 5294: 5293: 5282: 5276: 5275: 5273: 5264: 5258: 5257: 5247: 5223: 5217: 5216: 5200: 5190: 5184: 5183: 5155: 5149: 5148: 5140: 5134: 5125: 5119: 5114: 5108: 5107: 5105: 5104: 5098: 5091: 5083: 5077: 5076: 5074: 5062: 5053: 5048: 5042: 5039: 5033: 5032: 5019: 4994: 4988: 4982: 4976: 4975: 4968: 4959: 4958: 4951: 4942: 4941: 4938:The Toronto Star 4935: 4927: 4921: 4913: 4886: 4866: 4853: 4847: 4846: 4815: 4809: 4788:Isaiah Lankham, 4786: 4769: 4766: 4760: 4741: 4479:'s c. 1882 book 4439:base 60 examples 4435: 4410: 4406: 4402: 4398: 4386: 4382: 4378: 4374: 4366: 4328:Sun Microsystems 4214: 4196: 4181:William M. Kahan 4115: 4111: 4107: 4103: 4099: 4095: 4091: 4087: 4083: 4079: 4075: 4050: 4046: 4042: 4038: 4034: 4030: 4026: 4022: 4018: 4014: 4010: 4006: 4002: 3998: 3994: 3990: 3986: 3982: 3952:When digitizing 3909:preferred values 3875: 3871: 3868:and denominator 3867: 3864:whose numerator 3863: 3849: 3836:round to nearest 3833: 3825: 3818: 3800: 3796: 3792: 3788: 3780: 3776: 3760: 3757:is greater than 3756: 3752: 3746: 3740: 3736: 3732: 3718:Whether a value 3714: 3713:log(180) = 2.255 3710: 3709:log(165) = 2.217 3706: 3705:log(150) = 2.176 3672: 3670: 3669: 3664: 3626: 3625: 3615: 3614: 3602: 3562: 3514: 3510: 3485: 3471: 3459: 3447: 3440: 3437:, the increment 3429: 3425: 3421: 3411: 3409: 3408: 3403: 3389: 3378: 3343: 3286: 3282: 2828: 2815: 2796: 2783: 2747: 2733:machine learning 2719: 2717: 2716: 2711: 2709: 2708: 2702: 2685: 2682: 2608: 2605: 2544: 2537: 2530: 2523: 2519: 2456: 2447: 2437: 2428: 2418: 2333: 2329: 2325: 2265: 2261: 2257: 2212: 2210: 2209: 2204: 2202: 2198: 2197: 2188: 2182: 2142: 2138: 2137: 2128: 2122: 2075: 2071: 2060: 2056: 2045: 2018: 2016: 2015: 2010: 2008: 2004: 2003: 1994: 1988: 1948: 1944: 1943: 1934: 1928: 1881: 1877: 1866: 1862: 1851: 1811: 1809: 1808: 1803: 1801: 1797: 1784: 1775: 1764: 1760: 1759: 1750: 1727: 1723: 1722: 1713: 1684: 1673: 1648:two's complement 1626: 1624: 1623: 1618: 1616: 1612: 1599: 1590: 1579: 1575: 1574: 1565: 1542: 1538: 1537: 1528: 1499: 1488: 1484: 1469:Rounding half up 1439:is exactly 0.5. 1438: 1434: 1430: 1411: 1409: 1408: 1403: 1401: 1400: 1383: 1357: 1335: 1331: 1330: 1290: 1286: 1240: 1236: 1232: 1228: 1205: 1203: 1202: 1197: 1195: 1194: 1177: 1151: 1129: 1125: 1124: 1084: 1080: 1016: 1012: 1008: 1004: 1000: 971: 969: 968: 963: 961: 957: 939: 894: 890: 861: 859: 858: 853: 851: 847: 829: 806: 773: 769: 743: 735: 712: 708: 704: 700: 696: 681: 677: 644: 642: 640: 639: 634: 632: 631: 625: 616: 610: 601: 595: 586: 583: 582: 544: 539: 519: 514: 494: 474: 445:Approximating a 433: 410: 296:wavy equals sign 217: 209: 198: 190: 186: 159: 155: 151: 132: 125: 121: 118: 112: 110: 69: 45: 37: 21: 6191: 6190: 6186: 6185: 6184: 6182: 6181: 6180: 6151: 6150: 6120: 6119: 6116: 6111: 6110: 6104:Wayback Machine 6094: 6090: 6078: 6074: 6073: 6069: 6027: 6022: 6021: 6017: 6008: 6006: 6002: 5995: 5991: 5990: 5986: 5977: 5975: 5965: 5964: 5960: 5939: 5938: 5934: 5909: 5908: 5904: 5894: 5884: 5883: 5879: 5866:. ASTM. 2013 . 5862: 5852: 5840: 5829: 5828: 5824: 5817: 5813: 5800: 5799: 5795: 5786: 5784: 5780: 5779: 5775: 5763: 5759: 5746: 5745: 5741: 5721: 5720: 5716: 5707: 5705: 5694: 5693: 5686: 5675: 5671: 5658: 5657: 5653: 5646:ensl-00000007v2 5629:10.1.1.106.6652 5613: 5612: 5608: 5599: 5597: 5592: 5591: 5587: 5574: 5573: 5569: 5557: 5553: 5540: 5539: 5535: 5522: 5521: 5517: 5508: 5506: 5501: 5500: 5496: 5481: 5451:Revol, Nathalie 5448: 5447: 5443: 5434: 5432: 5424: 5423: 5419: 5406: 5405: 5401: 5392: 5390: 5363: 5355: 5354: 5350: 5341: 5339: 5310: 5305: 5304: 5300: 5291: 5289: 5284: 5283: 5279: 5271: 5266: 5265: 5261: 5225: 5224: 5220: 5213: 5192: 5191: 5187: 5157: 5156: 5152: 5142: 5141: 5137: 5126: 5122: 5115: 5111: 5102: 5100: 5096: 5089: 5085: 5084: 5080: 5064: 5063: 5056: 5049: 5045: 5040: 5036: 5020: 5013: 4996: 4995: 4991: 4983: 4979: 4970: 4969: 4962: 4953: 4952: 4945: 4933: 4928: 4919: 4914: 4887: 4883: 4855: 4854: 4850: 4817: 4816: 4812: 4787: 4783: 4778: 4773: 4772: 4767: 4763: 4742: 4738: 4733: 4673: 4656: 4650: 4637: 4629: 4539: 4473:human computers 4433: 4417: 4408: 4404: 4400: 4396: 4393: 4384: 4380: 4376: 4372: 4364: 4353:halting problem 4210: 4192: 4178: 4166:rounding to odd 4162:sticky rounding 4126: 4124:Double rounding 4113: 4109: 4105: 4101: 4097: 4093: 4089: 4085: 4081: 4077: 4073: 4070: 4057: 4048: 4044: 4040: 4036: 4032: 4028: 4024: 4020: 4016: 4012: 4008: 4004: 4000: 3996: 3992: 3988: 3984: 3980: 3962:error diffusion 3950: 3945: 3925: 3917:scaled rounding 3913:preferred value 3894: 3878:Farey sequences 3873: 3869: 3865: 3851: 3847: 3844: 3831: 3820: 3813: 3810:Scaled rounding 3798: 3794: 3790: 3786: 3778: 3774: 3767: 3758: 3754: 3748: 3742: 3738: 3734: 3719: 3712: 3708: 3704: 3682: 3680:Scaled rounding 3606: 3581: 3520: 3519: 3512: 3508: 3497: 3492: 3480: 3469: 3457: 3445: 3438: 3427: 3423: 3416: 3292: 3291: 3284: 3280: 3273: 3268: 2834: 2833: 2827: 2821: 2820: 2814: 2802: 2801: 2795: 2789: 2788: 2782: 2745: 2704: 2703: 2679: 2661: 2660: 2602: 2586: 2562: 2561: 2551: 2542: 2532: 2525: 2521: 2520:is 0.5, choose 2517: 2514: 2497: 2492: 2451: 2441: 2432: 2422: 2413: 2400:Double rounding 2365: 2331: 2327: 2323: 2316: 2263: 2259: 2255: 2244: 2172: 2168: 2164: 2112: 2111: 2107: 2081: 2080: 2073: 2062: 2058: 2047: 2043: 2032: 1978: 1974: 1970: 1918: 1917: 1913: 1887: 1886: 1879: 1868: 1864: 1853: 1849: 1834: 1772: 1768: 1738: 1734: 1704: 1700: 1689: 1688: 1675: 1671: 1660:round half down 1656: 1587: 1583: 1553: 1549: 1519: 1515: 1504: 1503: 1490: 1486: 1482: 1471: 1436: 1432: 1428: 1421: 1396: 1395: 1384: 1373: 1370: 1369: 1358: 1347: 1340: 1320: 1316: 1312: 1276: 1272: 1246: 1245: 1238: 1234: 1230: 1226: 1215: 1190: 1189: 1178: 1167: 1164: 1163: 1152: 1141: 1134: 1114: 1110: 1106: 1070: 1066: 1022: 1021: 1014: 1010: 1006: 1002: 998: 981: 950: 946: 929: 900: 899: 892: 888: 871: 840: 836: 819: 779: 778: 771: 767: 750: 741: 729: 710: 706: 705:is an integer, 702: 698: 694: 679: 675: 668: 659: 572: 571: 569: 542: 537: 517: 512: 492: 472: 447:rational number 431: 405: 305: 289:digital signals 263:ill-conditioned 223:round-off error 215: 207: 196: 188: 184: 157: 153: 149: 148:of the result, 133: 122: 116: 113: 70: 68: 58: 46: 35: 28: 23: 22: 15: 12: 11: 5: 6189: 6187: 6179: 6178: 6173: 6168: 6163: 6153: 6152: 6149: 6148: 6142: 6136: 6115: 6114:External links 6112: 6109: 6108: 6088: 6067: 6038:(3): 783–808. 6015: 5984: 5958: 5932: 5902: 5877: 5822: 5811: 5793: 5773: 5757: 5754:on 2021-02-25. 5739: 5714: 5684: 5669: 5666:on 2016-10-27. 5651: 5606: 5585: 5582:on 2021-03-01. 5567: 5551: 5548:on 2005-02-07. 5533: 5530:on 1999-05-08. 5515: 5494: 5479: 5461:(1 ed.). 5441: 5417: 5399: 5374:(4): 462–471. 5348: 5328:10.1.1.43.3309 5298: 5277: 5259: 5218: 5211: 5185: 5150: 5135: 5131:, 2002-01-21. 5120: 5109: 5078: 5054: 5043: 5034: 5011: 4989: 4977: 4960: 4943: 4881: 4848: 4829:(7): 610–621. 4810: 4794:Anne Schilling 4780: 4779: 4777: 4774: 4771: 4770: 4761: 4735: 4734: 4732: 4729: 4728: 4727: 4722: 4717: 4711: 4706: 4701: 4696: 4691: 4686: 4680: 4672: 4669: 4660:meteorologists 4652:Main article: 4649: 4646: 4636: 4633: 4628: 4625: 4538: 4535: 4533:was measured. 4416: 4413: 4392: 4389: 4345: 4344: 4338: 4331: 4325: 4318: 4307: 4245:transcendental 4219: 4200: 4177: 4174: 4125: 4122: 4069: 4066: 4056: 4053: 3949: 3946: 3944: 3941: 3924: 3923:Arbitrary bins 3921: 3907:When a set of 3893: 3890: 3843: 3840: 3784: 3766: 3763: 3681: 3678: 3674: 3673: 3662: 3659: 3656: 3653: 3650: 3647: 3644: 3641: 3638: 3635: 3632: 3629: 3624: 3621: 3618: 3613: 3609: 3605: 3601: 3598: 3595: 3592: 3589: 3584: 3580: 3577: 3574: 3571: 3568: 3565: 3561: 3558: 3555: 3552: 3549: 3546: 3543: 3540: 3537: 3534: 3531: 3528: 3496: 3493: 3491: 3488: 3450:numeral system 3444:The increment 3413: 3412: 3401: 3398: 3395: 3392: 3388: 3384: 3381: 3377: 3374: 3371: 3368: 3365: 3361: 3358: 3355: 3352: 3349: 3346: 3342: 3339: 3336: 3333: 3330: 3327: 3324: 3321: 3318: 3315: 3312: 3309: 3306: 3303: 3300: 3272: 3269: 3267: 3264: 3261: 3260: 3257: 3254: 3251: 3248: 3245: 3242: 3239: 3236: 3233: 3229: 3228: 3225: 3222: 3219: 3216: 3213: 3210: 3207: 3204: 3201: 3197: 3196: 3193: 3190: 3187: 3184: 3181: 3178: 3174: 3173: 3170: 3167: 3164: 3161: 3158: 3155: 3152: 3149: 3146: 3142: 3141: 3138: 3135: 3132: 3129: 3126: 3123: 3120: 3117: 3114: 3110: 3109: 3106: 3103: 3100: 3097: 3094: 3091: 3087: 3086: 3083: 3080: 3077: 3074: 3071: 3068: 3065: 3062: 3059: 3055: 3054: 3051: 3048: 3045: 3042: 3039: 3036: 3033: 3030: 3027: 3023: 3022: 3019: 3016: 3013: 3010: 3007: 3004: 3000: 2999: 2996: 2993: 2990: 2987: 2984: 2981: 2978: 2975: 2972: 2968: 2967: 2964: 2961: 2958: 2955: 2952: 2949: 2946: 2943: 2940: 2936: 2935: 2932: 2929: 2926: 2923: 2920: 2917: 2914: 2911: 2908: 2905: 2902: 2899: 2896: 2893: 2890: 2887: 2884: 2880: 2879: 2874: 2869: 2864: 2859: 2854: 2849: 2846: 2843: 2840: 2839:Half Toward 0 2837: 2831: 2830: 2824: 2818: 2817: 2811: 2808: 2805: 2799: 2798: 2792: 2786: 2785: 2778: 2777: 2774: 2771: 2768: 2765: 2762: 2758: 2757: 2754: 2751: 2744: 2741: 2721: 2720: 2707: 2701: 2698: 2695: 2692: 2689: 2680: 2678: 2675: 2672: 2669: 2666: 2663: 2662: 2659: 2656: 2653: 2650: 2647: 2644: 2641: 2638: 2635: 2632: 2629: 2626: 2623: 2620: 2617: 2614: 2611: 2603: 2601: 2598: 2595: 2592: 2591: 2589: 2584: 2581: 2578: 2575: 2572: 2569: 2550: 2547: 2513: 2510: 2496: 2493: 2491: 2488: 2473: 2472: 2469: 2466: 2459: 2458: 2449: 2439: 2430: 2420: 2396: 2395: 2392: 2389: 2386: 2383: 2380: 2364: 2361: 2354: 2340:United Kingdom 2315: 2312: 2279:Dutch rounding 2243: 2240: 2221:sign-magnitude 2214: 2213: 2201: 2194: 2191: 2185: 2181: 2178: 2175: 2171: 2167: 2163: 2160: 2157: 2154: 2151: 2148: 2145: 2141: 2134: 2131: 2125: 2121: 2118: 2115: 2110: 2106: 2103: 2100: 2097: 2094: 2091: 2088: 2031: 2028: 2020: 2019: 2007: 2000: 1997: 1991: 1987: 1984: 1981: 1977: 1973: 1969: 1966: 1963: 1960: 1957: 1954: 1951: 1947: 1940: 1937: 1931: 1927: 1924: 1921: 1916: 1912: 1909: 1906: 1903: 1900: 1897: 1894: 1833: 1830: 1813: 1812: 1800: 1796: 1793: 1790: 1787: 1781: 1778: 1771: 1767: 1763: 1756: 1753: 1747: 1744: 1741: 1737: 1733: 1730: 1726: 1719: 1716: 1710: 1707: 1703: 1699: 1696: 1655: 1652: 1628: 1627: 1615: 1611: 1608: 1605: 1602: 1596: 1593: 1586: 1582: 1578: 1571: 1568: 1562: 1559: 1556: 1552: 1548: 1545: 1541: 1534: 1531: 1525: 1522: 1518: 1514: 1511: 1470: 1467: 1463:floating-point 1420: 1417: 1413: 1412: 1399: 1394: 1391: 1388: 1385: 1382: 1379: 1376: 1372: 1371: 1368: 1365: 1362: 1359: 1356: 1353: 1350: 1346: 1345: 1343: 1338: 1334: 1329: 1326: 1323: 1319: 1315: 1311: 1308: 1305: 1302: 1299: 1296: 1293: 1289: 1285: 1282: 1279: 1275: 1271: 1268: 1265: 1262: 1259: 1256: 1253: 1214: 1211: 1207: 1206: 1193: 1188: 1185: 1182: 1179: 1176: 1173: 1170: 1166: 1165: 1162: 1159: 1156: 1153: 1150: 1147: 1144: 1140: 1139: 1137: 1132: 1128: 1123: 1120: 1117: 1113: 1109: 1105: 1102: 1099: 1096: 1093: 1090: 1087: 1083: 1079: 1076: 1073: 1069: 1065: 1062: 1059: 1056: 1053: 1050: 1047: 1044: 1041: 1038: 1035: 1032: 1029: 980: 977: 973: 972: 960: 956: 953: 949: 945: 942: 938: 935: 932: 928: 925: 922: 919: 916: 913: 910: 907: 870: 867: 863: 862: 850: 846: 843: 839: 835: 832: 828: 825: 822: 818: 815: 812: 809: 805: 802: 799: 796: 793: 789: 786: 749: 746: 736:refers to the 720:set up by the 667: 664: 658: 655: 652: 651: 648: 645: 630: 622: 619: 613: 607: 604: 598: 592: 589: 581: 567: 563: 562: 559: 556: 553: 549: 548: 545: 540: 535: 524: 523: 520: 515: 510: 502: 501: 498: 495: 490: 487:decimal number 482: 481: 479:decimal places 475: 470: 467: 465:decimal number 460: 459: 456: 453: 450: 442: 441: 438: 435: 428: 424: 423: 420: 417: 416:Example input 414: 404: 401: 378: 377: 374: 363: 338: 337: 336: 326: 299: 251:floating-point 152:, of rounding 135: 134: 49: 47: 40: 26: 24: 14: 13: 10: 9: 6: 4: 3: 2: 6188: 6177: 6174: 6172: 6169: 6167: 6164: 6162: 6159: 6158: 6156: 6146: 6143: 6140: 6137: 6132: 6131: 6126: 6123: 6118: 6117: 6113: 6105: 6101: 6098: 6092: 6089: 6084: 6077: 6071: 6068: 6063: 6059: 6054: 6049: 6045: 6041: 6037: 6033: 6026: 6019: 6016: 6001: 5994: 5988: 5985: 5973: 5969: 5962: 5959: 5955: 5952: 5945: 5944: 5936: 5933: 5929: 5927: 5924: 5921: 5915: 5914: 5906: 5903: 5899: 5890: 5889: 5881: 5878: 5873: 5869: 5865: 5859: 5855: 5848: 5844: 5836: 5832: 5826: 5823: 5820: 5815: 5812: 5807: 5806:libc.llvm.org 5803: 5797: 5794: 5783: 5777: 5774: 5770: 5766: 5761: 5758: 5753: 5749: 5743: 5740: 5735: 5731: 5727: 5726: 5718: 5715: 5704: 5700: 5699: 5691: 5689: 5685: 5682: 5678: 5673: 5670: 5665: 5661: 5655: 5652: 5647: 5643: 5639: 5635: 5630: 5625: 5622:(1): 85–102. 5621: 5617: 5610: 5607: 5595: 5589: 5586: 5581: 5577: 5571: 5568: 5564: 5560: 5555: 5552: 5547: 5543: 5537: 5534: 5529: 5525: 5519: 5516: 5504: 5498: 5495: 5490: 5486: 5482: 5476: 5472: 5468: 5464: 5460: 5456: 5452: 5445: 5442: 5431: 5427: 5421: 5418: 5413: 5409: 5403: 5400: 5389: 5385: 5381: 5377: 5373: 5369: 5362: 5358: 5357:Boldo, Sylvie 5352: 5349: 5338: 5334: 5329: 5324: 5320: 5316: 5309: 5302: 5299: 5288: 5281: 5278: 5270: 5263: 5260: 5255: 5251: 5246: 5241: 5237: 5233: 5229: 5222: 5219: 5214: 5212:0-8330-2601-1 5208: 5204: 5199: 5198: 5189: 5186: 5181: 5177: 5173: 5169: 5165: 5161: 5154: 5151: 5146: 5139: 5136: 5133: 5130: 5124: 5121: 5118: 5113: 5110: 5095: 5088: 5082: 5079: 5073: 5068: 5061: 5059: 5055: 5052: 5047: 5044: 5038: 5035: 5031: 5027: 5023: 5018: 5014: 5012:0-13-093904-8 5008: 5004: 5000: 4993: 4990: 4986: 4981: 4978: 4973: 4967: 4965: 4961: 4956: 4950: 4948: 4944: 4939: 4932: 4925: 4918: 4911: 4907: 4903: 4899: 4895: 4891: 4884: 4878: 4874: 4870: 4865: 4864: 4858: 4852: 4849: 4844: 4840: 4836: 4832: 4828: 4824: 4820: 4814: 4811: 4807: 4803: 4799: 4795: 4791: 4785: 4782: 4775: 4765: 4762: 4758: 4754: 4750: 4746: 4740: 4737: 4730: 4726: 4723: 4721: 4718: 4715: 4712: 4710: 4707: 4705: 4704:ISO/IEC 80000 4702: 4700: 4697: 4695: 4692: 4690: 4687: 4684: 4681: 4678: 4677:Cash rounding 4675: 4674: 4670: 4668: 4666: 4661: 4655: 4647: 4645: 4642: 4634: 4632: 4626: 4624: 4622: 4618: 4614: 4610: 4607:and in their 4606: 4602: 4598: 4595:In contrast, 4593: 4591: 4587: 4583: 4579: 4575: 4570: 4565: 4563: 4559: 4556: 4552: 4548: 4544: 4536: 4534: 4532: 4528: 4524: 4520: 4516: 4511: 4509: 4503: 4501: 4497: 4492: 4490: 4486: 4482: 4478: 4477:Simon Newcomb 4474: 4470: 4467: 4463: 4459: 4455: 4451: 4447: 4442: 4440: 4436: 4430: 4426: 4422: 4414: 4412: 4390: 4388: 4370: 4362: 4358: 4354: 4350: 4342: 4339: 4336: 4332: 4329: 4326: 4323: 4319: 4316: 4315:GNU C Library 4312: 4308: 4305: 4301: 4297: 4293: 4289: 4285: 4281: 4277: 4274:, written by 4273: 4269: 4265: 4262: 4261: 4260: 4258: 4254: 4249: 4246: 4242: 4238: 4233: 4231: 4227: 4221: 4217: 4213: 4208: 4204: 4198: 4195: 4188: 4186: 4182: 4173: 4171: 4167: 4163: 4157: 4155: 4151: 4146: 4145:IEEE 754-2008 4141: 4139: 4135: 4130: 4123: 4121: 4119: 4067: 4065: 4062: 4054: 4052: 3977: 3975: 3969: 3967: 3963: 3959: 3955: 3947: 3942: 3940: 3938: 3934: 3930: 3922: 3920: 3918: 3914: 3910: 3905: 3901: 3899: 3891: 3889: 3887: 3883: 3879: 3862: 3858: 3854: 3841: 3839: 3837: 3827: 3823: 3816: 3811: 3806: 3804: 3797:. The number 3782: 3777:into a value 3772: 3764: 3762: 3751: 3745: 3730: 3726: 3722: 3716: 3702: 3698: 3693: 3691: 3687: 3679: 3677: 3660: 3657: 3654: 3651: 3648: 3645: 3642: 3639: 3636: 3633: 3630: 3627: 3619: 3616: 3611: 3607: 3582: 3578: 3572: 3569: 3566: 3518: 3517: 3516: 3506: 3502: 3494: 3489: 3487: 3483: 3478: 3473: 3467: 3463: 3455: 3451: 3442: 3436: 3431: 3419: 3399: 3396: 3390: 3386: 3382: 3359: 3353: 3350: 3347: 3290: 3289: 3288: 3277: 3270: 3265: 3258: 3255: 3252: 3249: 3246: 3243: 3240: 3237: 3234: 3231: 3230: 3226: 3223: 3220: 3217: 3214: 3211: 3199: 3198: 3194: 3191: 3176: 3175: 3171: 3168: 3144: 3143: 3139: 3136: 3133: 3130: 3127: 3124: 3112: 3111: 3107: 3104: 3089: 3088: 3084: 3081: 3057: 3056: 3052: 3049: 3046: 3043: 3040: 3037: 3025: 3024: 3020: 3017: 3002: 3001: 2997: 2994: 2970: 2969: 2965: 2962: 2959: 2956: 2953: 2950: 2938: 2937: 2933: 2930: 2927: 2924: 2921: 2918: 2912: 2903: 2897: 2882: 2881: 2878: 2875: 2873: 2870: 2868: 2865: 2863: 2860: 2858: 2855: 2853: 2850: 2847: 2845:Half to Even 2844: 2841: 2838: 2825: 2812: 2809: 2806: 2793: 2780: 2779: 2759: 2748: 2742: 2740: 2738: 2734: 2729: 2724: 2696: 2690: 2687: 2676: 2673: 2667: 2657: 2654: 2651: 2648: 2642: 2636: 2627: 2621: 2618: 2612: 2609: 2596: 2587: 2582: 2576: 2570: 2567: 2560: 2559: 2558: 2556: 2548: 2546: 2539: 2535: 2528: 2511: 2509: 2505: 2503: 2494: 2489: 2487: 2485: 2481: 2476: 2470: 2467: 2464: 2463: 2462: 2454: 2450: 2445: 2440: 2435: 2431: 2426: 2421: 2416: 2412: 2411: 2410: 2408: 2403: 2401: 2393: 2390: 2387: 2384: 2381: 2378: 2377: 2376: 2372: 2370: 2362: 2360: 2358: 2353: 2350: 2348: 2343: 2341: 2335: 2326:is 0.5, then 2321: 2318:One may also 2313: 2311: 2308: 2306: 2301: 2299: 2294: 2292: 2288: 2284: 2280: 2276: 2272: 2267: 2258:is 0.5, then 2253: 2249: 2246:One may also 2241: 2239: 2235: 2233: 2228: 2226: 2222: 2217: 2199: 2192: 2189: 2183: 2179: 2176: 2173: 2169: 2165: 2158: 2152: 2149: 2146: 2143: 2139: 2132: 2129: 2123: 2119: 2116: 2113: 2108: 2101: 2095: 2092: 2089: 2086: 2079: 2078: 2077: 2076:is negative. 2069: 2065: 2054: 2050: 2041: 2037: 2034:One may also 2029: 2027: 2023: 2005: 1998: 1995: 1989: 1985: 1982: 1979: 1975: 1971: 1964: 1958: 1955: 1952: 1949: 1945: 1938: 1935: 1929: 1925: 1922: 1919: 1914: 1907: 1901: 1898: 1895: 1892: 1885: 1884: 1883: 1882:is negative. 1875: 1871: 1860: 1856: 1847: 1843: 1839: 1836:One may also 1831: 1829: 1827: 1823: 1822: 1816: 1798: 1791: 1788: 1779: 1776: 1769: 1765: 1761: 1754: 1751: 1745: 1742: 1739: 1735: 1731: 1728: 1724: 1717: 1714: 1708: 1705: 1701: 1697: 1694: 1687: 1686: 1685: 1682: 1678: 1669: 1668:round half up 1665: 1661: 1658:One may also 1653: 1651: 1649: 1644: 1642: 1638: 1637: 1631: 1613: 1606: 1603: 1594: 1591: 1584: 1580: 1576: 1569: 1566: 1560: 1557: 1554: 1550: 1546: 1543: 1539: 1532: 1529: 1523: 1520: 1516: 1512: 1509: 1502: 1501: 1500: 1497: 1493: 1480: 1476: 1475:round half up 1468: 1466: 1464: 1461:In practice, 1459: 1457: 1453: 1450:numbers with 1449: 1444: 1440: 1426: 1418: 1416: 1392: 1389: 1386: 1380: 1377: 1374: 1366: 1363: 1360: 1354: 1351: 1348: 1341: 1336: 1332: 1327: 1324: 1321: 1317: 1313: 1306: 1300: 1297: 1294: 1291: 1287: 1283: 1280: 1277: 1273: 1266: 1260: 1257: 1254: 1251: 1244: 1243: 1242: 1224: 1220: 1217:One may also 1212: 1210: 1186: 1183: 1180: 1174: 1171: 1168: 1160: 1157: 1154: 1148: 1145: 1142: 1135: 1130: 1126: 1121: 1118: 1115: 1111: 1107: 1100: 1094: 1091: 1088: 1085: 1081: 1077: 1074: 1071: 1067: 1060: 1054: 1051: 1048: 1042: 1036: 1033: 1030: 1027: 1020: 1019: 1018: 996: 992: 991: 986: 983:One may also 978: 976: 958: 954: 951: 947: 943: 940: 936: 933: 930: 926: 920: 914: 911: 908: 905: 898: 897: 896: 886: 882: 881: 877:(or take the 876: 873:One may also 868: 866: 848: 844: 841: 837: 833: 830: 826: 823: 820: 816: 810: 787: 784: 777: 776: 775: 765: 761: 760: 756:(or take the 755: 748:Rounding down 747: 745: 739: 738:sign function 733: 726: 723: 719: 714: 691: 689: 685: 673: 665: 663: 656: 649: 646: 620: 617: 611: 605: 602: 596: 590: 587: 568: 565: 564: 560: 557: 554: 551: 550: 546: 541: 536: 534: 530: 526: 525: 521: 516: 511: 508: 504: 503: 499: 496: 491: 488: 484: 483: 480: 476: 471: 468: 466: 462: 461: 457: 454: 451: 448: 444: 443: 439: 436: 434: 429: 426: 425: 421: 418: 415: 412: 411: 408: 402: 400: 398: 394: 390: 386: 381: 375: 372: 368: 364: 361: 360: 355: 352:, in general 351: 347: 343: 339: 334: 330: 329: 327: 324: 320: 319: 318: 315: 313: 309: 304: 303: 297: 292: 290: 286: 282: 277: 275: 271: 266: 264: 260: 256: 252: 248: 244: 240: 236: 232: 228: 224: 219: 213: 205: 200: 194: 182: 178: 174: 170: 163: 147: 143: 139: 131: 128: 120: 109: 106: 102: 99: 95: 92: 88: 85: 81: 78: – 77: 73: 72:Find sources: 66: 62: 56: 55: 50:This article 48: 44: 39: 38: 33: 19: 6128: 6095:OFCM, 2005: 6091: 6082: 6070: 6035: 6031: 6018: 6007:. Retrieved 5987: 5976:. Retrieved 5971: 5961: 5950: 5947: 5942: 5935: 5925: 5922: 5919: 5917: 5912: 5905: 5892: 5887: 5880: 5863: 5857: 5846: 5830: 5825: 5814: 5805: 5796: 5785:. Retrieved 5776: 5760: 5752:the original 5742: 5724: 5717: 5706:. Retrieved 5697: 5672: 5664:the original 5654: 5619: 5615: 5609: 5598:. Retrieved 5588: 5580:the original 5570: 5554: 5546:the original 5536: 5528:the original 5518: 5507:. Retrieved 5505:. 1999-04-18 5497: 5458: 5444: 5433:. Retrieved 5420: 5411: 5402: 5391:. Retrieved 5371: 5367: 5351: 5340:. Retrieved 5318: 5314: 5301: 5290:. Retrieved 5280: 5262: 5235: 5231: 5221: 5196: 5188: 5163: 5159: 5153: 5138: 5123: 5112: 5101:. Retrieved 5081: 5046: 5037: 5029: 5021: 5016: 4998: 4992: 4980: 4937: 4923: 4896:(1): 47–48. 4893: 4889: 4862: 4851: 4826: 4822: 4813: 4797: 4784: 4764: 4739: 4683:Data binning 4657: 4638: 4630: 4594: 4589: 4566: 4540: 4512: 4504: 4499: 4495: 4493: 4484: 4480: 4465: 4461: 4457: 4445: 4443: 4421:clay tablets 4418: 4394: 4375:=1+10 where 4359:is true but 4347:There exist 4346: 4310: 4271: 4267: 4259:(binary64): 4250: 4234: 4223: 4211: 4193: 4190: 4179: 4165: 4161: 4158: 4142: 4137: 4133: 4131: 4127: 4071: 4058: 3978: 3970: 3951: 3936: 3929:data binning 3926: 3906: 3902: 3895: 3860: 3856: 3852: 3845: 3835: 3828: 3821: 3814: 3807: 3768: 3749: 3743: 3728: 3724: 3720: 3717: 3694: 3685: 3683: 3675: 3504: 3500: 3498: 3481: 3474: 3443: 3432: 3417: 3414: 3278: 3274: 2848:Half to Odd 2810:Away From 0 2725: 2722: 2552: 2540: 2533: 2526: 2515: 2506: 2498: 2477: 2474: 2460: 2452: 2443: 2433: 2424: 2414: 2404: 2397: 2373: 2366: 2346: 2344: 2336: 2319: 2317: 2309: 2302: 2295: 2290: 2286: 2282: 2278: 2274: 2270: 2268: 2251: 2247: 2245: 2236: 2231: 2229: 2218: 2215: 2067: 2063: 2052: 2048: 2039: 2035: 2033: 2024: 2021: 1873: 1869: 1858: 1854: 1845: 1841: 1837: 1835: 1825: 1824:rather than 1819: 1817: 1814: 1680: 1676: 1667: 1663: 1659: 1657: 1645: 1640: 1639:rather than 1634: 1632: 1629: 1495: 1491: 1478: 1474: 1472: 1460: 1445: 1441: 1424: 1422: 1414: 1241:(included). 1233:) such that 1222: 1218: 1216: 1208: 994: 988: 984: 982: 974: 884: 878: 874: 872: 864: 763: 757: 753: 751: 731: 727: 715: 692: 671: 669: 660: 406: 382: 379: 357: 316: 307: 300: 293: 281:quantization 278: 267: 239:square roots 220: 201: 173:rounding off 172: 168: 167: 138: 123: 117:October 2017 114: 104: 97: 90: 83: 71: 59:Please help 54:verification 51: 5412:gcc.gnu.org 5166:: 506–507. 5026:p. 165 4694:Guard digit 4519:age heaping 4429:reciprocals 4425:Mesopotamia 4276:Abraham Ziv 4118:slide rules 3783:significant 3755:165 = 27225 2776:Stochastic 2502:random seed 1448:fixed-point 869:Rounding up 181:approximate 6161:Arithmetic 6155:Categories 6125:"Rounding" 6009:2011-08-19 5978:2014-01-30 5787:2022-08-30 5708:2022-08-30 5600:2022-07-18 5509:2022-08-29 5489:2009939668 5463:Birkhäuser 5435:2008-11-14 5393:2016-08-02 5342:2016-08-02 5292:2021-09-12 5103:2010-11-24 5072:1502.02551 5003:p. 85 4776:References 4725:Truncation 4621:ECMAScript 4617:JavaScript 4523:ABCC Index 4515:Jörg Baten 4452:Z25.1 and 4361:unprovable 4300:Windows NT 4235:Using the 3733:rounds to 3701:E12 series 3456:(that is, 2555:stochastic 754:round down 647:{0, 0, 1} 452:399 / 941 385:idempotent 367:symmetries 344:will be a 259:accumulate 243:logarithms 87:newspapers 76:"Rounding" 6130:MathWorld 6053:10230/481 5898:as minus. 5624:CiteSeerX 5323:CiteSeerX 4957:. Oracle. 4808:, p. 186. 4590:precision 4578:overflows 4569:libraries 4423:found in 4218:correctly 3958:dithering 3896:Finished 3658:≠ 3617:⁡ 3397:× 2829:(toward + 2816:(toward − 2813:Half Down 2807:Toward 0 2797:(toward + 2784:(toward − 2737:dithering 2700:⌋ 2694:⌊ 2691:− 2671:⌋ 2665:⌊ 2649:− 2646:⌋ 2640:⌊ 2631:⌋ 2625:⌊ 2622:− 2613:− 2600:⌋ 2594:⌊ 2571:⁡ 2442:2.5 < 2423:2.0 < 2184:− 2170:− 2153:⁡ 2147:− 2096:⁡ 1976:− 1959:⁡ 1953:− 1930:− 1902:⁡ 1795:⌉ 1786:⌈ 1740:− 1732:− 1709:− 1610:⌋ 1601:⌊ 1561:− 1555:− 1547:− 1364:≥ 1318:− 1301:⁡ 1295:− 1261:⁡ 1158:≥ 1112:− 1095:⁡ 1089:− 1055:⁡ 1037:⁡ 952:− 944:− 915:⁡ 842:− 834:− 543:3.01 × 10 538:300999999 389:monotonic 314:in 1892. 6100:Archived 6062:35494384 6000:Archived 5274:. Intel. 5254:14829295 5180:26299526 5094:Archived 4859:(2002). 4843:35883481 4671:See also 4558:variable 4531:literacy 4527:numeracy 4322:ENS Lyon 4253:GNU MPFR 4226:IEEE 754 4154:strictfp 3937:relative 3505:multiple 3420:= 2.1784 2728:function 2446:< 3.0 2427:< 2.5 2352:denormal 2298:IEEE 754 2200:⌉ 2166:⌈ 2140:⌋ 2109:⌊ 2006:⌋ 1972:⌊ 1946:⌉ 1915:⌈ 1799:⌋ 1770:⌊ 1762:⌋ 1736:⌊ 1725:⌉ 1702:⌈ 1614:⌉ 1585:⌈ 1577:⌉ 1551:⌈ 1540:⌋ 1517:⌊ 1473:One may 1381:⌋ 1375:⌊ 1355:⌉ 1349:⌈ 1333:⌋ 1314:⌊ 1288:⌉ 1274:⌈ 1175:⌉ 1169:⌈ 1149:⌋ 1143:⌊ 1127:⌉ 1108:⌈ 1082:⌋ 1068:⌊ 1034:truncate 990:truncate 959:⌋ 948:⌊ 937:⌉ 931:⌈ 875:round up 849:⌉ 838:⌈ 827:⌋ 821:⌊ 752:One may 709:is just 354:discrete 323:function 237:such as 212:accurate 193:fraction 179:with an 169:Rounding 5874:. E-29. 5559:mathlib 5388:1850330 5203:255–293 4910:2691491 4665:Celsius 4555:integer 4547:FORTRAN 4415:History 4292:Solaris 4280:PowerPC 3212:−1.495 3125:−0.495 3038:+0.505 2951:+1.505 2872:Average 2862:Average 2852:Average 2826:Half Up 2484:pSeries 2480:zSeries 880:ceiling 643:⁠ 570:⁠ 529:integer 507:integer 419:Result 348:of the 227:integer 187:with $ 185:23.4476 101:scholar 6060: 5926:number 5769:GitHub 5765:libmcr 5681:GitHub 5677:crlibm 5644: 5626: 5563:GitHub 5487: 5477: 5386: 5325: 5252: 5209: 5178: 5009: 4908: 4879: 4841: 4804: 4751:, see 4586:printf 4335:x86-64 4309:IBM's 4047:, and 4031:, and 4011:, and 4001:0.9677 3993:0.3091 3991:, and 3989:0.7451 3985:0.9204 3981:0.9677 3898:lumber 3884:, and 3880:, the 3824:> 1 2750:Value 531:using 493:2.1784 473:1.6667 469:5 / 3 455:3 / 7 350:domain 346:subset 333:metric 245:, and 216:123500 208:123456 191:, the 177:number 146:Graphs 103: 96: 89: 82: 74: 6079:(PDF) 6058:S2CID 6028:(PDF) 6003:(PDF) 5996:(PDF) 5858:Power 5384:S2CID 5364:(PDF) 5311:(PDF) 5272:(PDF) 5250:S2CID 5097:(PDF) 5090:(PDF) 5067:arXiv 4934:(PDF) 4920:(PDF) 4906:JSTOR 4839:S2CID 4731:Notes 4658:Some 4562:array 4541:Most 4517:used 4288:SPARC 4199:every 4100:. If 3501:power 3462:power 3259:0.04 3256:−1.8 3232:−1.8 3227:0.05 3224:−1.5 3221:0.05 3218:−1.5 3200:−1.5 3195:0.04 3192:−1.2 3177:−1.2 3172:0.04 3169:−0.8 3145:−0.8 3140:0.05 3137:−0.5 3134:0.05 3131:−0.5 3113:−0.5 3108:0.04 3105:−0.2 3090:−0.2 3085:0.04 3082:+0.2 3058:+0.2 3053:0.05 3050:+0.5 3047:0.05 3044:+0.5 3026:+0.5 3021:0.04 3018:+0.8 3003:+0.8 2998:0.04 2995:+1.2 2971:+1.2 2966:0.05 2963:+1.5 2960:0.05 2957:+1.5 2939:+1.5 2934:0.04 2931:+1.8 2883:+1.8 2568:Round 2536:− 0.5 2529:+ 0.5 2455:= 3.0 2436:= 2.5 2417:= 2.0 2357:radix 2289:, or 2070:− 0.5 2055:+ 0.5 1876:+ 0.5 1861:− 0.5 1683:− 0.5 1498:+ 0.5 993:, or 883:, or 762:, or 759:floor 718:index 555:48.2 518:23200 513:23217 497:2.18 437:22/7 393:order 342:range 247:sines 197:1.414 189:23.45 108:JSTOR 94:books 5954:any. 5923:even 5485:LCCN 5475:ISBN 5207:ISBN 5176:PMID 5007:ISBN 4877:ISBN 4827:C-26 4802:ISBN 4641:U.S. 4599:and 4549:and 4460:and 4454:ASTM 4444:The 4341:LLVM 4294:and 4268:ml4j 4239:and 4224:The 4136:and 4037:0.97 4033:2.94 4029:2.63 4025:1.89 4021:0.97 4017:0.01 3997:0.01 3803:base 3711:and 3646:> 3634:> 2781:Down 2531:and 2482:and 2038:(or 1840:(or 1662:(or 1477:(or 1390:< 1221:(or 1184:< 987:(or 912:ceil 730:sgn( 371:bias 162:SMIL 80:news 6048:hdl 6040:doi 5868:doi 5767:on 5730:doi 5679:on 5642:HAL 5634:doi 5561:on 5467:doi 5376:doi 5333:doi 5240:doi 5168:doi 5164:115 4898:doi 4869:doi 4831:doi 4757:VAT 4619:or 4609:IDL 4605:DOM 4601:SVG 4597:CSS 4582:PHP 4296:x86 4284:AIX 4266:'s 4264:IBM 4207:ulp 4150:x87 4132:In 4088:of 3817:= 1 3769:In 3737:or 3723:∈ ( 3608:log 3484:= 1 3250:−2 3244:−2 3241:−2 3238:−2 3235:−2 3209:−2 3206:−2 3203:−2 3189:−2 3186:−1 3183:−1 3180:−2 3163:−1 3157:−1 3154:−1 3151:−1 3148:−1 3122:−1 3119:−1 3116:−1 3102:−1 3099:−1 3093:−1 3015:+1 3009:+1 2989:+1 2983:+1 2980:+1 2977:+1 2974:+1 2948:+1 2945:+1 2942:+1 2925:+2 2919:+2 2916:+1 2913:+2 2910:+2 2907:+2 2904:+2 2901:+2 2898:+2 2895:+2 2892:+1 2889:+2 2886:+1 2252:and 2150:sgn 2093:sgn 2072:if 2057:if 1956:sgn 1899:sgn 1878:if 1863:if 1298:sgn 1258:sgn 1225:): 1092:sgn 1052:sgn 997:): 887:): 766:): 693:If 558:45 276:". 229:or 218:". 171:or 63:by 6157:: 6127:. 6081:. 6056:. 6046:. 6036:69 6034:. 6030:. 5998:. 5856:. 5845:. 5833:. 5804:. 5701:. 5687:^ 5640:. 5632:. 5620:41 5618:. 5483:. 5473:. 5465:. 5457:. 5428:. 5410:. 5382:. 5372:57 5370:. 5366:. 5331:. 5321:. 5319:47 5317:. 5313:. 5248:. 5236:30 5234:. 5230:. 5205:. 5174:. 5162:. 5092:. 5057:^ 5028:. 5015:. 5005:. 4963:^ 4946:^ 4936:. 4922:. 4904:. 4894:73 4892:. 4875:. 4837:. 4825:. 4796:: 4792:, 4510:. 4441:. 4367:, 4286:, 4187:: 4172:. 4051:. 4043:, 4039:, 4027:, 4023:, 4019:: 4007:, 4003:, 3987:, 3983:, 3888:. 3855:= 3826:. 3761:. 3750:ab 3727:, 3707:, 3430:. 3253:0 3247:0 3215:0 3166:0 3160:0 3128:0 3096:0 3079:0 3076:0 3073:0 3070:0 3067:0 3064:0 3061:0 3041:0 3035:0 3032:0 3029:0 3012:0 3006:0 2992:0 2986:0 2954:0 2928:0 2922:0 2877:SD 2867:SD 2857:SD 2794:Up 2739:. 2486:. 2293:. 2285:, 2281:, 2277:, 2273:, 2066:= 2051:= 1872:= 1857:= 1828:. 1679:= 1643:. 1494:= 895:. 774:. 744:. 713:. 684:+∞ 621:12 606:12 591:12 477:4 399:. 306:, 294:A 291:. 241:, 199:. 6133:. 6085:. 6064:. 6050:: 6042:: 6012:. 5981:. 5928:. 5895:n 5870:: 5808:. 5790:. 5771:. 5736:. 5732:: 5711:. 5648:. 5636:: 5603:. 5565:. 5512:. 5491:. 5469:: 5438:. 5414:. 5396:. 5378:: 5345:. 5335:: 5295:. 5256:. 5242:: 5215:. 5182:. 5170:: 5106:. 5075:. 5069:: 4940:. 4926:. 4912:. 4900:: 4885:. 4871:: 4845:. 4833:: 4551:C 4434:π 4409:π 4405:π 4401:π 4397:π 4385:k 4381:n 4377:k 4373:n 4365:n 4298:/ 4290:/ 4282:/ 4212:y 4194:y 4114:n 4110:y 4106:x 4102:n 4098:y 4094:x 4090:z 4086:x 4082:z 4078:n 4074:n 3874:m 3870:n 3866:m 3861:n 3859:/ 3857:m 3853:y 3848:x 3832:x 3822:b 3815:s 3799:m 3795:x 3791:m 3787:y 3779:y 3775:x 3744:x 3739:b 3735:a 3731:) 3729:b 3725:a 3721:x 3661:1 3655:b 3652:, 3649:0 3643:b 3640:, 3637:0 3631:x 3628:, 3623:) 3620:x 3612:b 3604:( 3600:d 3597:n 3594:u 3591:o 3588:r 3583:b 3579:= 3576:) 3573:b 3570:, 3567:x 3564:( 3560:r 3557:e 3554:w 3551:o 3548:P 3545:o 3542:T 3539:d 3536:n 3533:u 3530:o 3527:r 3513:b 3509:x 3482:m 3470:m 3468:( 3458:m 3446:m 3439:m 3418:x 3400:m 3394:) 3391:m 3387:/ 3383:x 3380:( 3376:d 3373:n 3370:u 3367:o 3364:r 3360:= 3357:) 3354:m 3351:, 3348:x 3345:( 3341:e 3338:l 3335:p 3332:i 3329:t 3326:l 3323:u 3320:M 3317:o 3314:T 3311:d 3308:n 3305:u 3302:o 3299:r 3285:m 3281:x 2835:) 2832:∞ 2822:) 2819:∞ 2803:) 2800:∞ 2790:) 2787:∞ 2697:x 2688:x 2677:1 2674:+ 2668:x 2658:1 2655:+ 2652:x 2643:x 2637:= 2634:) 2628:x 2619:x 2616:( 2610:1 2597:x 2588:{ 2583:= 2580:) 2577:x 2574:( 2543:y 2534:x 2527:x 2522:y 2518:x 2453:x 2444:x 2434:x 2425:x 2415:x 2332:x 2328:y 2324:x 2264:x 2260:y 2256:x 2193:2 2190:1 2180:| 2177:x 2174:| 2162:) 2159:x 2156:( 2144:= 2133:2 2130:1 2124:+ 2120:| 2117:x 2114:| 2105:) 2102:x 2099:( 2090:= 2087:y 2074:x 2068:x 2064:y 2059:x 2053:x 2049:y 2044:x 1999:2 1996:1 1990:+ 1986:| 1983:x 1980:| 1968:) 1965:x 1962:( 1950:= 1939:2 1936:1 1926:| 1923:x 1920:| 1911:) 1908:x 1905:( 1896:= 1893:y 1880:x 1874:x 1870:y 1865:x 1859:x 1855:y 1850:x 1792:x 1789:2 1780:2 1777:1 1766:= 1755:2 1752:1 1746:+ 1743:x 1729:= 1718:2 1715:1 1706:x 1698:= 1695:y 1681:x 1677:y 1672:x 1607:x 1604:2 1595:2 1592:1 1581:= 1570:2 1567:1 1558:x 1544:= 1533:2 1530:1 1524:+ 1521:x 1513:= 1510:y 1496:x 1492:y 1487:x 1483:x 1437:x 1433:x 1429:x 1393:0 1387:x 1378:x 1367:0 1361:x 1352:x 1342:{ 1337:= 1328:| 1325:x 1322:| 1310:) 1307:x 1304:( 1292:= 1284:| 1281:x 1278:| 1270:) 1267:x 1264:( 1255:= 1252:y 1239:y 1235:x 1231:x 1227:y 1187:0 1181:x 1172:x 1161:0 1155:x 1146:x 1136:{ 1131:= 1122:| 1119:x 1116:| 1104:) 1101:x 1098:( 1086:= 1078:| 1075:x 1072:| 1064:) 1061:x 1058:( 1049:= 1046:) 1043:x 1040:( 1031:= 1028:y 1015:x 1011:y 1007:x 1003:x 999:y 955:x 941:= 934:x 927:= 924:) 921:x 918:( 909:= 906:y 893:x 889:y 845:x 831:= 824:x 817:= 814:) 811:x 808:( 804:r 801:o 798:o 795:l 792:f 788:= 785:y 772:x 768:y 742:x 734:) 732:x 711:x 707:y 703:x 699:x 695:x 680:y 676:x 629:} 618:5 612:, 603:4 597:, 588:3 580:{ 432:π 373:. 362:. 359:Z 335:. 302:≈ 298:( 158:y 154:x 150:y 130:) 124:( 119:) 115:( 105:· 98:· 91:· 84:· 57:. 34:. 20:)

Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.