Python Decimal Module
The decimal module provides support for fast correctly-rounded decimal floating point arithmetic. It offers several advantages over the float datatype, such as exact representation of decimal numbers (no binary floating-point errors) and user-definable precision.
Importing the Module
from decimal import Decimal, getcontextCreating Decimals
You can create Decimal instances from integers, strings, floats, or tuples.
From Integers and Strings
Creating from integers or strings is exact.
from decimal import Decimal
d1 = Decimal(10)
d2 = Decimal("10.5")
print(d1) # Output: 10
print(d2) # Output: 10.5From Floats
Creating from floats can reveal the inherent inaccuracy of binary floating-point numbers.
d3 = Decimal(0.1)
print(d3)
# Output: 0.1000000000000000055511151231257827021181583404541015625To get the exact decimal representation of a float as a human would write it, convert it to a string first.
d4 = Decimal(str(0.1))
print(d4) # Output: 0.1Arithmetic Operations
Decimal objects support standard arithmetic operations.
a = Decimal("0.1")
b = Decimal("0.2")
c = a + b
print(c) # Output: 0.3
# Compare with float: 0.1 + 0.2 = 0.30000000000000004Context and Precision
The decimal module uses a context to control precision and rounding.
Getting and Setting Precision
You can access the current context using getcontext().
from decimal import getcontext
# Default precision is usually 28
print(getcontext().prec)
# Change precision
getcontext().prec = 6
print(Decimal(1) / Decimal(7)) # Output: 0.142857Rounding
You can also set the rounding mode.
import decimal
getcontext().rounding = decimal.ROUND_HALF_UP
print(Decimal("2.5").quantize(Decimal("1"))) # Output: 3
getcontext().rounding = decimal.ROUND_HALF_DOWN
print(Decimal("2.5").quantize(Decimal("1"))) # Output: 2Comparison
Decimal objects can be compared with other Decimal objects and integers. Comparison with floats is also supported but can be tricky due to precision.
print(Decimal("0.1") == 0.1) # False (because float 0.1 is not exactly 0.1)
print(Decimal("0.5") == 0.5) # True (0.5 is exactly representable in binary)