[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

*To*: hal*Subject*: another dumb question about numbers*From*: Kent M Pitman <KMP@stony-brook.scrc.symbolics.com>*Date*: Fri, 3 Nov 89 19:21 EST*Cc*: rrrs-authors*In-Reply-To*: <8911022257.AA13507@zurich.ai.mit.edu>

Date: Thu, 2 Nov 89 17:57:49 est From: Hal Abelson <hal@zurich.ai.mit.edu> Why not have round, floor, ceiling, and truncate produce exact integers, and provide some other names for the analogous (more primitive?) operations that do not force exactness? This doesn't seem right. What if I ask How far is it from Boston to San Francisco and someone returns replies "About 3000 miles." If I do (FLOOR 3000) to get an integer, that oughtn't make the number exact. All it should gurantee me is that the result has no fractional part (which in this case was also true of the input, but only incidentally). It's a common thing for humans to not specify a fractional part of something if the number is no good in the ones-ies place, but there's no necessary reason. e.g., a number like 3.5 might really be 3.5+/-2.1 because of other errors introduced in the computation. So to round that number and say that it is an exact 3 or exact 4 neglects that an exact 2 or even an exact 1 might be the right exact answer if all the accumulated error was in that direction.

**References**:**another dumb question about numbers***From:*hal (Hal Abelson)

- Prev by Date:
**floor & inexact->exact** - Next by Date:
**Re: floor & inexact->exact** - Prev by thread:
**another dumb question about numbers** - Next by thread:
**floor & inexact->exact** - Index(es):