2
$\begingroup$

Mathematica doesn't want to solve my exact system with 3 equations and 3 variables.

AdjustPar20[x_] := a x^b Exp[c x]
Solve[AdjustPar20[0.05] == 9.37126 && AdjustPar20'[0.05] == 0 && 
  AdjustPar20[0] == 0, {a, b, c}]

As an explanations why this cannot be solved the program replies:

{}
Solve::ifun: Inverse functions are being used by Solve, so some solutions may not be found; 
             use Reduce for complete solution information. >>

Any ideas?

$\endgroup$
5
$\begingroup$

Solve[] works much better with exact arithmetic:

AdjustPar20[x_] := a x^b Exp[c x]
p = 937126/100000;
sol = Solve[
           AdjustPar20[5/100] == p && 
           ((D[AdjustPar20[t], t] == 0) /. t -> 5/100) &&
           AdjustPar20[0] == 0, {a, b, c}, Reals]

ParametricPlot3D[{a, b, c} /. sol, {a, p, 10 p},
                 PlotRange -> {Automatic, {-1, 1}, Automatic}, 
                 AxesLabel -> {a, b, c}, BoxRatios -> 1]

enter image description here

Please note that AdjustPar20[0] == 0 isn't adding any information since your selected function form already ensures it

$\endgroup$
  • 1
    $\begingroup$ Note that AdjustPar20[0] == 0 isn't adding information since your selected function form already ensures it. $\endgroup$ – Dr. belisarius Apr 29 '13 at 12:46
  • $\begingroup$ How does one know to solve with these circumstances (rational numbers, real domain)? $\endgroup$ – BoLe Apr 29 '13 at 12:51
  • $\begingroup$ @BoLe, in general, inexact numbers and Solve[] don't mix too well... $\endgroup$ – J. M. will be back soon Apr 29 '13 at 12:57
  • $\begingroup$ @ belisarius thanks for you suggestion, but my version of Mathematica, 8, cannot solve this. Solve::ivar: 298.48501198004965` is not a valid variable. >> Solve[2.687648609756841*10^-437 a == 9.37126 && 1.007466030596621*10^-433 a == 0 && 3.211110349737091*10^-722 a == 0.01, {a, 298.485, -2221.2}, Reals] Do you think I should define one more equation(instead of the 3rd one), in order to get one exact value of a? $\endgroup$ – energyMax Apr 29 '13 at 13:08
  • 1
    $\begingroup$ @Gasper Yes, you need one more equation. But we can't help with that since the equation should represent your problem conditions ... and we ignore them- $\endgroup$ – Dr. belisarius Apr 29 '13 at 13:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.