What is the difference between rounding Decimals with quantize vs the built in round function?

user_5689745 0
Viewed 377

When working with the built in decimal module in python I can round decimals as follows.

Decimal(50.212345).quantize(Decimal('0.01'))
> Decimal('50.21')

But I can also round the same number with the built in round function

round(Decimal(50.212345), 2)
> Decimal('50.21')

Why would I use one instead of the other when rounding Decimals? In previous answers about rounding decimals, users suggested to use quantize because the built in round function would return a value of type float. Based on my testing, these both return a Decimal. Other than syntax, is there a reason to choose one over the other?

decimal floating-point rounding python-3.x python
1 Answers

The return types aren't always the same. round() used with a single argument actually returns an int:

>>> round(5.3)
5
>>> round(decimal.Decimal("5.3"))
5

Other than that, suit yourself. quantize() is especially handy if you want a deoimal rounded to "the same" precision as another decimal you already have.

>>> x = decimal.Decimal("123.456")
>>> x*x
Decimal('15241.383936')
>>> (x*x).quantize(x)
Decimal('15241.384')

See? The code doing this doesn't have to know that x originally had 3 digits after the decimal point. Just passing x to quantize() forces the function to round back to the same precision as the original x, regardless of what that may be.

quantize() is also necessary if you want to use a rounding mode other than the default nearest/even.

>>> (x*x).quantize(x, decimal.ROUND_FLOOR)
Decimal('15241.383')
user_27055420
Related