[Maxima-bugs] [ maxima-Bugs-1729432 ] onep1 / FIX

[SourceForge.net <no...@so...>]

Bugs item #1729432, was opened at 2007-06-01 07:32
Message generated for change (Comment added) made by willisbl
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=1729432&group_id=4933

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Lisp Core
Group: Includes proposed fix
>Status: Closed
>Resolution: Fixed
Priority: 4
Private: No
Submitted By: Barton Willis (willisbl)
Assigned to: Nobody/Anonymous (nobody)
Summary: onep1  / FIX

Initial Comment:
The function onep1 compares a big float to bigfloatone.
But the value of bigfloatone depends on fpprec; so

(%i1) fpprec : 20$
(%i2) x : 1.0b0$
(%i3) ?onep1(x);
(%o3) true
(%i4) fpprec : 21$
(%i5) ?onep1(x);
(%o5) false

The function zerop1 doesn't have this problem (but
zerop1 shouldn't have call alike1).

Also, I think that ?onep1(rat(1)) should be true.


----------------------------------------------------------------------

>Comment By: Barton Willis (willisbl)
Date: 2007-06-04 12:23

Message:
Logged In: YES 
user_id=895922
Originator: YES

Suggest fix applied -- see simp.lisp CVS revision 1.39

----------------------------------------------------------------------

Comment By: Barton Willis (willisbl)
Date: 2007-06-01 22:14

Message:
Logged In: YES 
user_id=895922
Originator: YES

Proposed code:

(defun $bfloatp (x) 
  (and (consp x)
       (consp (car x))
       (eq (caar x) 'bigfloat)))

(defun zerop1 (x)
  (or
   (and (integerp x) (= 0 x))
   (and (floatp x) (= 0.0 x))
   (and ($bfloatp x) (= 0 (second x)))))

(defun onep1 (x)
  (or
   (and (integerp x) (= 1 x))
   (and (floatp x) (= 1.0 x))
   (and ($bfloatp x) (zerop1 (sub x 1)))))

The new bfloat code is needed because onep1 sometimes
recieves an invalid expressions (from taylor, perhaps).


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=1729432&group_id=4933