On 4/2/07, Kostas Oikonomou <ko at research.att.com> wrote:
> I've reduced the linsolve issue to this example:
>
> linsolve([x=2,y=3],[x])
Looks like the error occurs as SOLVEX (src/solve.lisp) attempts
to output an error message about inconsistent equations.
I think TFGELI is Gaussian elimination but I don't know how
its return value is supposed to be interpreted.
Maxima 5.11.0cvs http://maxima.sourceforge.net
Using Lisp CLISP 2.38 (2006-01-24)
(%i1) :lisp (trace solvex merror tfgeli)
(%i1) linsolve([x=26,y=38],[x]);
1. Trace: (SOLVEX '(((MPLUS SIMP) -26 $X) ((MPLUS SIMP) -38 $Y)) '($X)
'NIL 'NIL)
2. Trace: (TFGELI 'XA* '2 '2)
2. Trace: TFGELI ==> (NIL (2) (1))
2. Trace: (MERROR '"Inconsistent equations: ~A" '(2))
Maxima encountered a Lisp error:
CAR: 2 is not a list
Automatically continuing.
To reenable the Lisp debugger set *debugger-hook* to nil.
Solving for both variables is successful:
(%i2) linsolve([x=26,y=38],[x, y]);
1. Trace: (SOLVEX '(((MPLUS SIMP) -26 $X) ((MPLUS SIMP) -38 $Y)) '($X
$Y) 'NIL 'NIL)
2. Trace: (TFGELI 'XA* '2 '3)
2. Trace: TFGELI ==> (NIL NIL (1 2))
1. Trace: SOLVEX ==> ((MLIST) ((MEQUAL) $X 26) ((MEQUAL) $Y 38))
(%o2) [x = 26, y = 38]
FWIW
Robert