15. Арифметика з плаваючою комою: проблеми та обмеження

Floating-point numbers are represented in computer hardware as base 2 (binary) fractions. For example, the decimal fraction 0.625 has value 6/10 + 2/100 + 5/1000, and in the same way the binary fraction 0.101 has value 1/2 + 0/4 + 1/8. These two fractions have identical values, the only real difference being that the first is written in base 10 fractional notation, and the second in base 2.

На жаль, більшість десяткових дробів не можна представити точно як двійкові дроби. Наслідком цього є те, що загалом десяткові числа з плаваючою комою, які ви вводите, є лише наближеними до двійкових чисел з плаваючою комою, які фактично зберігаються в машині.

Задачу легше зрозуміти спочатку з основою 10. Розглянемо дріб 1/3. Ви можете наблизити це як дріб з основою 10:

0.3

або, краще,

0.33

або, краще,

0.333

і так далі. Незалежно від того, скільки цифр ви готові записати, результат ніколи не буде рівно 1/3, а буде дедалі кращим наближенням 1/3.

Таким же чином, незалежно від того, скільки цифр з основою 2 ви бажаєте використовувати, десяткове значення 0,1 не може бути представлено точно як дріб з основою 2. За основою 2 1/10 — нескінченно повторюваний дріб

0.0001100110011001100110011001100110011001100110011...

Зупиніться на будь-якій кінцевій кількості бітів, і ви отримаєте наближення. На більшості сучасних машин числа з плаваючою точкою наближаються за допомогою двійкового дробу з чисельником, використовуючи перші 53 біти, починаючи зі старшого біта, а знаменник – ступінь двійки. У випадку 1/10 двійковий дріб дорівнює 3602879701896397 / 2 ** 55, що близько до справжнього значення 1/10, але не зовсім дорівнює йому.

Many users are not aware of the approximation because of the way values are displayed. Python only prints a decimal approximation to the true decimal value of the binary approximation stored by the machine. On most machines, if Python were to print the true decimal value of the binary approximation stored for 0.1, it would have to display:

>>> 0.1
0.1000000000000000055511151231257827021181583404541015625

That is more digits than most people find useful, so Python keeps the number of digits manageable by displaying a rounded value instead:

>>> 1 / 10
0.1

Просто пам’ятайте, навіть якщо надрукований результат виглядає як точне значення 1/10, фактичне збережене значення є найближчим двійковим дрібом, який можна представити.

Цікаво, що існує багато різних десяткових чисел, які мають однаковий найближчий наближений двійковий дріб. Наприклад, числа 0,1 і 0,100000000000000001 і 0,100000000000000005511151231257827021181583404541015625 наближено до 360287970751596 Оскільки всі ці десяткові значення мають однакове наближення, будь-яке з них може бути відображено, зберігаючи інваріант eval(repr(x)) == x.

Історично склалося так, що підказка Python і вбудована функція repr() вибирали одну з 17 значущих цифр, 0.10000000000000001. Починаючи з Python 3.1, Python (у більшості систем) тепер може вибирати найкоротший із них і просто відображати 0.1.

Зауважте, що це в самій природі двійкового числа з плаваючою комою: це не помилка в Python і не є помилкою у вашому коді. Ви побачите те саме на всіх мовах, які підтримують арифметику з плаваючою комою вашого апаратного забезпечення (хоча деякі мови можуть не відображати різницю за замовчуванням або в усіх режимах виводу).

For more pleasant output, you may wish to use string formatting to produce a limited number of significant digits:

>>> format(math.pi, '.12g')  # give 12 significant digits
'3.14159265359'

>>> format(math.pi, '.2f')   # give 2 digits after the point
'3.14'

>>> repr(math.pi)
'3.141592653589793'

Важливо усвідомлювати, що насправді це ілюзія: ви просто округлюєте відображення справжнього значення машини.

One illusion may beget another. For example, since 0.1 is not exactly 1/10, summing three values of 0.1 may not yield exactly 0.3, either:

>>> 0.1 + 0.1 + 0.1 == 0.3
False

Also, since the 0.1 cannot get any closer to the exact value of 1/10 and 0.3 cannot get any closer to the exact value of 3/10, then pre-rounding with round() function cannot help:

>>> round(0.1, 1) + round(0.1, 1) + round(0.1, 1) == round(0.3, 1)
False

Though the numbers cannot be made closer to their intended exact values, the math.isclose() function can be useful for comparing inexact values:

>>> math.isclose(0.1 + 0.1 + 0.1, 0.3)
True

Alternatively, the round() function can be used to compare rough approximations:

>>> round(math.pi, ndigits=2) == round(22 / 7, ndigits=2)
True

Binary floating-point arithmetic holds many surprises like this. The problem with «0.1» is explained in precise detail below, in the «Representation Error» section. See Examples of Floating Point Problems for a pleasant summary of how binary floating-point works and the kinds of problems commonly encountered in practice. Also see The Perils of Floating Point for a more complete account of other common surprises.

Як сказано ближче до кінця, «простих відповідей немає». Тим не менш, не будьте надмірно обережні з плаваючою комою! Помилки в операціях з плаваючою комою в Python успадковуються від апаратного забезпечення з плаваючою комою, і на більшості машин вони становлять не більше 1 частини з 2**53 на операцію. Цього більш ніж достатньо для більшості завдань, але ви повинні мати на увазі, що це не десяткова арифметика і що кожна операція з плаваючою точкою може зазнати нової помилки округлення.

Хоча патологічні випадки дійсно існують, для більшості випадкового використання арифметики з плаваючою комою ви побачите очікуваний результат, якщо просто округлите відображення кінцевих результатів до очікуваної кількості десяткових цифр. str() зазвичай достатньо, і для більш точного контролю дивіться специфікатори формату методу str.format() у Синтаксис рядка формату.

Для випадків використання, які вимагають точного десяткового представлення, спробуйте використовувати модуль decimal, який реалізує десяткову арифметику, придатну для програм бухгалтерського обліку та програм високої точності.

Інша форма точної арифметики підтримується модулем fractions, який реалізує арифметику на основі раціональних чисел (таким чином числа, такі як 1/3, можуть бути представлені точно).

If you are a heavy user of floating-point operations you should take a look at the NumPy package and many other packages for mathematical and statistical operations supplied by the SciPy project. See <https://scipy.org>.

Python provides tools that may help on those rare occasions when you really do want to know the exact value of a float. The float.as_integer_ratio() method expresses the value of a float as a fraction:

>>> x = 3.14159
>>> x.as_integer_ratio()
(3537115888337719, 1125899906842624)

Since the ratio is exact, it can be used to losslessly recreate the original value:

>>> x == 3537115888337719 / 1125899906842624
True

The float.hex() method expresses a float in hexadecimal (base 16), again giving the exact value stored by your computer:

>>> x.hex()
'0x1.921f9f01b866ep+1'

This precise hexadecimal representation can be used to reconstruct the float value exactly:

>>> x == float.fromhex('0x1.921f9f01b866ep+1')
True

Оскільки представлення є точним, воно корисне для надійного перенесення значень між різними версіями Python (незалежність від платформи) та обміну даними з іншими мовами, які підтримують той самий формат (такими як Java і C99).

Another helpful tool is the sum() function which helps mitigate loss-of-precision during summation. It uses extended precision for intermediate rounding steps as values are added onto a running total. That can make a difference in overall accuracy so that the errors do not accumulate to the point where they affect the final total:

>>> 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 == 1.0
False
>>> sum([0.1] * 10) == 1.0
True

The math.fsum() goes further and tracks all of the «lost digits» as values are added onto a running total so that the result has only a single rounding. This is slower than sum() but will be more accurate in uncommon cases where large magnitude inputs mostly cancel each other out leaving a final sum near zero:

>>> arr = [-0.10430216751806065, -266310978.67179024, 143401161448607.16,
...        -143401161400469.7, 266262841.31058735, -0.003244936839808227]
>>> float(sum(map(Fraction, arr)))   # Exact summation with single rounding
8.042173697819788e-13
>>> math.fsum(arr)                   # Single rounding
8.042173697819788e-13
>>> sum(arr)                         # Multiple roundings in extended precision
8.042178034628478e-13
>>> total = 0.0
>>> for x in arr:
...     total += x                   # Multiple roundings in standard precision
...
>>> total                            # Straight addition has no correct digits!
-0.0051575902860057365

15.1. Помилка представлення

У цьому розділі детально пояснюється приклад «0.1» і показано, як ви можете самостійно виконати точний аналіз подібних випадків. Передбачається базове знайомство з двійковим представленням із плаваючою комою.

Representation error стосується того факту, що деякі (насправді більшість) десяткових дробів не можуть бути представлені точно як двійкові (за основою 2) дроби. Це головна причина, чому Python (або Perl, C, C++, Java, Fortran та багато інших) часто не відображає точне десяткове число, яке ви очікуєте.

Why is that? 1/10 is not exactly representable as a binary fraction. Since at least 2000, almost all machines use IEEE 754 binary floating-point arithmetic, and almost all platforms map Python floats to IEEE 754 binary64 «double precision» values. IEEE 754 binary64 values contain 53 bits of precision, so on input the computer strives to convert 0.1 to the closest fraction it can of the form J/2**N where J is an integer containing exactly 53 bits. Rewriting

1 / 10 ~= J / (2**N)

як

J ~= 2**N / 10

and recalling that J has exactly 53 bits (is >= 2**52 but < 2**53), the best value for N is 56:

>>> 2**52 <=  2**56 // 10  < 2**53
True

That is, 56 is the only value for N that leaves J with exactly 53 bits. The best possible value for J is then that quotient rounded:

>>> q, r = divmod(2**56, 10)
>>> r
6

Since the remainder is more than half of 10, the best approximation is obtained by rounding up:

>>> q+1
7205759403792794

Therefore the best possible approximation to 1/10 in IEEE 754 double precision is:

7205759403792794 / 2 ** 56

Ділення чисельника та знаменника на два скорочує дріб до:

3602879701896397 / 2 ** 55

Зауважте, що оскільки ми округлили в більшу сторону, це насправді трохи більше, ніж 1/10; якби ми не округляли в більшу сторону, частка була б трохи меншою за 1/10. Але ні в якому разі не може бути рівно 1/10!

So the computer never «sees» 1/10: what it sees is the exact fraction given above, the best IEEE 754 double approximation it can get:

>>> 0.1 * 2 ** 55
3602879701896397.0

If we multiply that fraction by 10**55, we can see the value out to 55 decimal digits:

>>> 3602879701896397 * 10 ** 55 // 2 ** 55
1000000000000000055511151231257827021181583404541015625

meaning that the exact number stored in the computer is equal to the decimal value 0.1000000000000000055511151231257827021181583404541015625. Instead of displaying the full decimal value, many languages (including older versions of Python), round the result to 17 significant digits:

>>> format(0.1, '.17f')
'0.10000000000000001'

The fractions and decimal modules make these calculations easy:

>>> from decimal import Decimal
>>> from fractions import Fraction

>>> Fraction.from_float(0.1)
Fraction(3602879701896397, 36028797018963968)

>>> (0.1).as_integer_ratio()
(3602879701896397, 36028797018963968)

>>> Decimal.from_float(0.1)
Decimal('0.1000000000000000055511151231257827021181583404541015625')

>>> format(Decimal.from_float(0.1), '.17')
'0.10000000000000001'