Comparing functional paradigms for exact real arithmetic

There are two standard approaches to exact real number computation within functional programming. Under one approach, real numbers are represented using an ordinary data type (for example, integer streams), and operations on them are implemented as functions on this type. In the second approach, the representation is hidden by encapsulating real numbers within an abstract data type.

In the talk, I will first give an overview of the two approaches, and then discuss preliminary results from a general investigation relating the two.