Exact real arithmetic for Tcl
نویسنده
چکیده
Once in a great while, the programmer has need for floating point arithmetic in greater precision than the hardware provides. This capability is usually not performance critical, since most performance critical code deals with measured quantities in the physical world, where it is hard to imagine a measurement precision that exceeds the sixteen or so decimal digits provided by standard IEEE-754 double precision arithmetic. Instead, the high-precision arithmetic is used to guard against catastrophic loss of significance, where a difference is computed between floating point numbers of nearly the same magnitude, or against accumulation of roundoff errors. In the course of developing numeric software, the analyst will usually rectify these problems as a matter of course, but the need arises to verify that nothing has been missed, and to compute intermediate constants (such as coefficients of power series) so that they are guaranteed accurate.
منابع مشابه
A Tcl-based Self-configuring Embedded System Debugger
The Tcl Environment for Extensible Modeling is a software system from Bell Labs for the simulation, hardware emulation and debugging of heterogeneous multiprocessor embedded systems. These embedded systems contain one or more digital signal processors or microcontrollers that execute real-time software written in assembly language and C. Tcl provides an environment in which embedded system desi...
متن کاملExact Real Arithmetic with Automatic Error
The common approach to real arithmetic on computers is floating point arithmetic, which can produce erroneous results due to roundoff errors. An alternative is exact real arithmetic and in this project such arithmetic is implemented in the well-known computer system Mathematica by the use of constructive real numbers. All basic operations are implemented as well as the common elementary functio...
متن کاملFrom Coinductive Proofs to Exact Real Arithmetic
We give a coinductive characterization of the set of continuous functions defined on a compact real interval, and extract certified programs that construct and combine exact real number algorithms with respect to the binary signed digit representation of real numbers. The data type corresponding to the coinductive definition of continuous functions consists of finitely branching non-wellfounded...
متن کاملRealLib: an Efficient Implementation of Exact Real Arithmetic
This paper is an introduction to the RealLib package for exact real number computations. The library provides certified accuracy, but tries to achieve this at performance close to the performance of hardware floating point for problems that do not require higher precision. The paper gives the motivation and features of the design of the library and compares it to other packages for exact real a...
متن کامل