Generalised floating point addition: Difference between revisions

IfIt is possible, defineto addition forimplement floating point numbersarithmetic wherein many different ways; though the digitsstandard areIEEE storedarithmetic implementation on modern hardware usually uses a fixed-width binary implementation (using a total of 32 bits for single-precision floats — <code>float</code> in ana arbitrarynumber base.of languages eg.— theand digits64 bits for double-precision floats — <code>double</code>) many other schemes can be storedused. asFor binaryexample, the base need not be binary (alternatives include decimal, [[wp:Binary-coded decimal|binary-coded decimal]], orand even [[balanced ternary]]) and there is no requirement that a fixed number of digits be used.

ImplementYou should implement the code in a generalised form (such as a [[wp:Template (programming)|Template]], [[wp:Modular programming|Module]] or [[wp:Mixin|Mixin]] etc) that permits reusing of the code for different [[wp:Base_(exponentiation)#In_numeral_systems|Bases]].

If it is not possible to implement code inby providing an implementation that uses the ''syntax'' of the specific language then:

* noteNote the reason why this is.

* performDemonstrate that the language still supports the semantics of generalised floating point by implementing the ''test case'' using a built-in code or externala library.

Use the Template to defineDefine [[wp:Arbitrary-precision arithmetic|Arbitraryarbitrary precision addition]] on floating point numbers stored(e.g., inencoding those numbers using Binary Coded Decimal, or using an arbitrary precision integer arithmetic library). Calculate the terms for -7 to 21 in this sequence of calculations:

PerformYou will either need to implement multiplication, or perform the multiplication ofby 81 by using repeated additions. The results will always be 1e72.