I'm using decimal module to avoid float rounding errors. In my case the values are money so I want two decimal places.
To do that I do decimal.getcontext().prec = 2 but then I get some surprising results which make me think I'm missing something. In this code the first assertion works, but the second fails
from decimal import getcontext, Decimal
assert Decimal("3000") + Decimal("20") == 3020
getcontext().prec = 2
assert Decimal("3000") + Decimal("20") == 3020 # fails
Since 3000 and 20 are integers I was expecting this to hold, but I get 3000 instead. Any ideas on what is happening?
decimal does not implement fixed-point arithmetic directly. It implements base 10 floating-point arithmetic. The precision (prec) is the total number of significant digits retained and has nothing to do with the position of the radix point.
Try displaying the computed value in your last example:
>>> Decimal("3000") + Decimal("20")
Decimal('3.0E+3')
The exact result (3020) is rounded back to the 2 (because you set prec to 2) most significant digits, so the trailing "20" is thrown away.
If, e.g., you want 2 places after the decimal point, you'll have to arrange for that yourself. Search the docs for the question "Once I have valid two place inputs, how do I maintain that invariant throughout an application?".