Fixed Decimals (fixeddecimal)

To the best of my knowledge R does not have a package that implements (exact) decimal arithmetic. This package aims at providing some of this functionality for simple tasks. I have found exact decimals useful for some simple/short calculations. For example, calculations in a report where a reviewer might check a calculation with a few decimal digits by hand and will expect a certain result (which cannot be guaranteed with binary floating point numbers).

The decimal representation in this package is built on the data type bigz provided by the gmp package.

All operations (e.g. +, -, *, /, mean) on decimals x and y will require the decimals to have the same number of fractional decimal digits and will retain the number of decimals digits of the input variables. Operations *, / and mean will round the result (away from zero). Multiple operations of this type will lead to multiple rounding (try to use prod).

Implemented

• Arithmetic operations (+, -, *, /)
• sum, prod, mean
• min, max, abs
• relational operators (==, <, ...)
• round.decimal
• Utilities (c.decimal, print.decimal etc.)