Bugs item #760722, was opened at 20030625 13:37
Message generated for change (Comment added) made by robert_dodier
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=760722&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  Integration
Group: None
>Status: Closed
>Resolution: Works For Me
Priority: 5
Submitted By: Nobody/Anonymous (nobody)
Assigned to: Nobody/Anonymous (nobody)
Summary: errors in integrate and factor
Initial Comment:
2 6 4 4 6 2 8
432 a y  504 a y + 6 a y  3 a

2 2 11/2
2 (y + a )
(C26) integrate(%,y,minf,inf);
Typeerror in KERNEL::OBJECTNOTFIXNUMERRORHANDLER:
NIL is not of type FIXNUM
The same error occured in FACTOR on
sqrt(y^2+a^2)*(432*a^2*y^6504*a^4*y^4+6*a^6*y^23*a^8)/(2*y^12+12*a^2*y^10+30*a^4*y^8+30*a^8*y^2
+12*a^10*y^2+2*a^12)
These errors occured in the following version:
Maxima version: 5.9.0
Maxima build date: 21:17 2/9/2003
host type: i386redhatlinuxgnu
lispimplementationtype: CMU Common Lisp
lispimplementationversion: 18d

>Comment By: Robert Dodier (robert_dodier)
Date: 20060409 22:58
Message:
Logged In: YES
user_id=501686
I don't see the " NIL is not of type FIXNUM" errors cited by
the original poster.
(1) (432 * a^2 * y^6  504 * a^4 * y^4 + 6 * a^6 * y^2  3 *
a^8) / (2*(y^2 + a^2)^(11/2));
integrate(%,y,minf,inf); => 0 after spurious asksign for y
and (meaningful) asksign for a.
From what I can tell, 0 is correct: realroots applied the
numerator of the above expr yields [y = 36136957/33554432,y
= 36136957/33554432], and integrating from 0 to
36136957/33554432 and from 36136957/33554432 to inf shows
the pieces are equal and opposite. Also quad_qagi (quadpack
function for infinite interval) seems to show integral is 0
to within tolerance.
(By the way, I got the above expression by inspecting the
HTML page source  the ascii art layout is comprehensible
that way.)
(2) sqrt(y^2+a^2) *
(432*a^2*y^6504*a^4*y^4+6*a^6*y^23*a^8) /
(2*y^12+12*a^2*y^10+30*a^4*y^8+30*a^8*y^2 + 12*a^10*y^2+2*a^12);
factor(%); => pretty much the same expression (no error)
Works the same with 5.9.1 / CMUCL 19a and 5.9.3 / Clisp 2.34
(both Linux).
Closing this report as "works for me" (but I'll open another
about the spurious asksign for y).

Comment By: Stavros Macrakis (macrakis)
Date: 20030706 22:05
Message:
Logged In: YES
user_id=588346
Please use the linear form of expressions  this bug reporting
system loses formatting. I believe the original expression was
(432*a^2*y^2504*a^4*y^4+6*a^6*y^23*a^9)
/
(2 * (y^2 + a^2) ^(11/2) )
Is that right?
In 5.9.0/gcl/Windows, I don't get the error you report. ON
the other hand, Maxima stupidly asks whether y is pos or neg
(but y only appears as y^2). It does give a result in terms of
a after a few questions. I have no idea whether this result is
correct.

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=760722&group_id=4933
