Rounding decimals with new Python format function
Question:
How do I round a decimal to a particular number of decimal places using the Python 3.0 format function?
Answers:
To round x to n decimal places use:
"{0:.{1}f}".format(x,n)
where 0 and 1 stand for the first and second arguments of the str.format() method, respectively.
Here’s a typical, useful example…:
>>> n = 4
>>> p = math.pi
>>> '{0:.{1}f}'.format(p, n)
'3.1416'
the nested {1} takes the second argument, the current value of n, and applies it as specified (here, to the “precision” part of the format — number of digits after the decimal point), and the outer resulting {0:.4f} then applies. Of course, you can hardcode the 4 (or whatever number of digits) if you wish, but the key point is, you don’t have to!
Even better…:
>>> '{number:.{digits}f}'.format(number=p, digits=n)
'3.1416'
…instead of the murky “argument numbers” such as 0 and 1 above, you can choose to use shiny-clear argument names, and pass the corresponding values as keyword (aka “named“) arguments to format — that can be so much more readable, as you see!!!
In Python 3.x a format string contains replacement fields indicated by braces thus::
".... {0: format_spec} ....".format(value)
The format spec has the general layout:
[[fill]align][sign][pad][width][,][.precision][type]
So, for example leaving out all else but width, precision and type code, a decimal or floating point number could be formatted as:
>>>print("The value of pi is {0:10.7f} to 7 decimal places.".format(math.pi))
This would print as:
The value of pi is 3.1415927 to 7 decimal places.
I just found out that it is possible to combine both the {0} and the {digits} notation. This is especially useful when you want to round all variables to a pre-specified number of decimals with 1 declaration:
sName = 'Nander'
fFirstFloat = 1.12345
fSecondFloat = 2.34567
fThirdFloat = 34.5678
dNumDecimals = 2
print( '{0} found the following floats: {1:.{digits}f}, {2:.{digits}f}, {3:.{digits}f}'.format(sName, fFirstFloat, fSecondFloat, fThirdFloat, digits=dNumDecimals))
# Nander found the following floats: 1.12, 2.35, 34.57
=============================================================
EDIT:
Since all answers (including mine) where using the ‘old’ {0}.format() method, I’m now adding the new f-strings method (Formatted String Literals, Python 3.6+ I think) too:
x = 1.2345
y = 23.45678
print(f'x = {x:.2f} and y = {y:.2f}')
#x = 1.23 and y = 23.46
An updated answer based on [Alex Martelli]’s solution but using Python 3.6.2 and it’s updated format syntax I would suggest:
>>> n=4
>>> p=math.pi
>>> f'{p:.{n}f}'
'3.1416'
But by choosing your variables wisely your code becomes self documenting
>>> precision = 4
>>> pi = math.pi
>>> f'{pi:.{precision}f}'
'3.1416'
How do I round a decimal to a particular number of decimal places using the Python 3.0 format function?
To round x to n decimal places use:
"{0:.{1}f}".format(x,n)
where 0 and 1 stand for the first and second arguments of the str.format() method, respectively.
Here’s a typical, useful example…:
>>> n = 4
>>> p = math.pi
>>> '{0:.{1}f}'.format(p, n)
'3.1416'
the nested {1} takes the second argument, the current value of n, and applies it as specified (here, to the “precision” part of the format — number of digits after the decimal point), and the outer resulting {0:.4f} then applies. Of course, you can hardcode the 4 (or whatever number of digits) if you wish, but the key point is, you don’t have to!
Even better…:
>>> '{number:.{digits}f}'.format(number=p, digits=n)
'3.1416'
…instead of the murky “argument numbers” such as 0 and 1 above, you can choose to use shiny-clear argument names, and pass the corresponding values as keyword (aka “named“) arguments to format — that can be so much more readable, as you see!!!
In Python 3.x a format string contains replacement fields indicated by braces thus::
".... {0: format_spec} ....".format(value)
The format spec has the general layout:
[[fill]align][sign][pad][width][,][.precision][type]
So, for example leaving out all else but width, precision and type code, a decimal or floating point number could be formatted as:
>>>print("The value of pi is {0:10.7f} to 7 decimal places.".format(math.pi))
This would print as:
The value of pi is 3.1415927 to 7 decimal places.
I just found out that it is possible to combine both the {0} and the {digits} notation. This is especially useful when you want to round all variables to a pre-specified number of decimals with 1 declaration:
sName = 'Nander'
fFirstFloat = 1.12345
fSecondFloat = 2.34567
fThirdFloat = 34.5678
dNumDecimals = 2
print( '{0} found the following floats: {1:.{digits}f}, {2:.{digits}f}, {3:.{digits}f}'.format(sName, fFirstFloat, fSecondFloat, fThirdFloat, digits=dNumDecimals))
# Nander found the following floats: 1.12, 2.35, 34.57
=============================================================
EDIT:
Since all answers (including mine) where using the ‘old’ {0}.format() method, I’m now adding the new f-strings method (Formatted String Literals, Python 3.6+ I think) too:
x = 1.2345
y = 23.45678
print(f'x = {x:.2f} and y = {y:.2f}')
#x = 1.23 and y = 23.46
An updated answer based on [Alex Martelli]’s solution but using Python 3.6.2 and it’s updated format syntax I would suggest:
>>> n=4
>>> p=math.pi
>>> f'{p:.{n}f}'
'3.1416'
But by choosing your variables wisely your code becomes self documenting
>>> precision = 4
>>> pi = math.pi
>>> f'{pi:.{precision}f}'
'3.1416'