Thanks Emmanuel for your precious answer. But It generates some few new
questions :
- is there a place in the documentation where I can find the information on
`solve()` and mainly its options ?- if I understand clearly z_{6497} is an
integer but how to fix it to zero --- when the number change at each iteration
- sympy seems to be the good approach but it is not self evident that to call y
one must typpeset sol2[0][x]- the giac way is certainly the better but it keeps
no track of the variable's order.I have tried my solution assuming l>0 on the 3
conditions but it changes nothing.----- Mail d’origine -----
De: Emmanuel Charpentier <emanuel.charpent...@gmail.com>
À: sage-support <sage-support@googlegroups.com>
Envoyé: Mon, 29 Nov 2021 11:03:37 +0100 (CET)
Objet: [sage-support] Re: Constrained optimization with strange result.
Variables of the form z_xxxx are integer variables created by Maxima, which
attempts to give you also the complex roots, if any, thus ignoring the
assumptions on x, y and l. Note that :sage: solve(FOC[0], x)
---------------------------------------------------------------------------
[ Snip… ]TypeError: Computation failed since Maxima requested additional
constraints; using the 'assume' command before evaluation *may* help (example
of legal syntax is 'assume(l>0)', see `assume?` for more details)
Is l positive, negative or zero?
sage: with assuming(l>0): print(solve(FOC[0], x))
[
x == (l*p_x)^(1/a)
]
sage: with assuming(l<0): print(solve(FOC[0], x))
[
x^a == l*p_x
]
sage: with assuming(l<0): print(solve(FOC[0], x, to_poly_solve=True))
[x == (l*p_x)^(1/a)*e^(2*I*pi*z4353/a)]
Interestingly:sage: solve([FOC[0]==0,FOC[1]==0,FOC[2]==0],x,y,l)
[[l == (1/p_y), x == (p_x/p_y)^(1/a)*e^(2*I*pi*z3540/a), y ==
-(p_x*(p_x/p_y)^(1/a)*e^(2*I*pi*z3540/a) - R)/p_y]]
sage: solve([FOC[0]==0,FOC[1]==0,FOC[2]==0],x,y,l, algorithm="sympy")
[{x: (p_x/p_y)^(1/a), l: 1/p_y, y: -(p_x*(p_x/p_y)^(1/a) - R)/p_y}]
sage: solve([FOC[0]==0,FOC[1]==0,FOC[2]==0],x,y,l, algorithm="fricas")
[[l == (1/p_y), x == (p_x/p_y)^(1/a)*e^(2*I*pi*z3891/a), y ==
-(p_x*(p_x/p_y)^(1/a)*e^(2*I*pi*z3891/a) - R)/p_y]]
sage: giac.solve(giac(FOC),giac([x,y,l])).sage()
[[(p_x/p_y)^(1/a), -(p_x*(p_x/p_y)^(1/a) - R)/p_y, 1/p_y]]
HTH,Le dimanche 28 novembre 2021 à 22:13:12 UTC+1, cyrille piatecki a écrit :
On my computer the solution of
var('a x y p_x p_y D Rev R l')
assume(a,'real')
assume(x,'real')
assume(y,'real')
assume(p_x,'real')
assume(p_y,'real')
assume(D,'real')
assume(Rev,'real')
assume(R,'real')
assume(l,'real')
assume(a<1)
assume(a>0)
assume(p_x>0)
assume(p_y>0)
assume(R>0)
U =(1/(a+1))*x^(a+1)+y
show(LatexExpr(r'\text{La fonction d}^\prime\text{utilité est }U(x,y) = '),U)
D= x*p_x + y*p_y
show(LatexExpr(r'\text{La Dépense } D = '),D)
Rev= R
show(LatexExpr(r'\text{Le Revenu } Rev = '),R)
L=U+l*(Rev-D)
show(LatexExpr(r'\text{Le lagrangien est } \mathcal{L}(x, y, λ) = '),L)
FOC = [diff(L,x),diff(L,y),diff(L,l)]
show(LatexExpr(r'\text{Les condition du premier ordre sont }
\left\{\begin{array}{c}\mathcal{L}_x= 0\\\mathcal{L}_y= 0\\\mathcal{L}_λ=
0\end{array}\right. '))
show(LatexExpr(r'\text{soit }'))
show(LatexExpr(r'\mathcal{L}_x= 0 \Longleftrightarrow '),FOC[0]==0)
show(LatexExpr(r'\mathcal{L}_y= 0 \Longleftrightarrow '),FOC[1]==0)
show(LatexExpr(r'\mathcal{L}_λ= 0 \Longleftrightarrow '),FOC[2]==0)
sol=solve([FOC[0]==0,FOC[1]==0,FOC[2]==0],x,y,l)
show(sol)
Is nearly correct, but an extra complex exponential term multiplies $x$ and
then modifies $y$. Even as an element I do not understand its form :
$e^{(2iπz_{5797}a)}$
Could some one explain why ?
--
You received this message because you are subscribed to the Google Groups
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-support+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/sage-support/e745f8a7-df76-4def-b3b9-09ac7684e9a3n%40googlegroups.com.
--
You received this message because you are subscribed to the Google Groups
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-support+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/sage-support/1541168298.2230365.1638199996780.JavaMail.zimbra%40univ-orleans.fr.
