0
$\begingroup$

I have a function which has one variable $r$ and two parameters $M,\lambda$. Whether or not the function has roots, depend on the parameters $M,\lambda$. I want to obtain a parameter space such that the function doesn't have roots. Naively, if the function is positive (or negative but this won't happen in this context) for all $r$ for a given set of parameters I know that there will be no roots. I want to check this condition for a range of parameter values, and therefore obtain a parameter plot. For example, if the function had been quadratic, I could simply check the determinant. But this is not a polynomial function, and has fractional powers in $r$. I am providing the functions here (the precision is set at 10 digits).

f[r_] := (r^2 (-1.000000000 + 2.000000000^(-1.000000000 Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]) (r/m)^Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]) (-1.000000000 (-1.000000000 + Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)])^2 + 2.000000000^(-1.000000000 Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]) (1.000000000 + Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)])^2 (r/m)^Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]))/(-1.000000000 + Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)] + 2.000000000^(-1.000000000 Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]) (1.000000000 + Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]) (r/m)^Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)])^2 + m^2 \[Lambda]^2

Physically, the $0 < \lambda < 1$ and $M > 1$. Ideally, I would like a RegionPlot to obtain the parameter space. I am just not sure how to operationally impose the condition that the function is always positive. So far, I tried

Reduce[ForAll[r,f[r] > 0],r]

but this doesn't give me anything useful. In fact, if the plot is obtained, I would not need an analytical expression for the condition.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

If you redefine f to include m and lambda

Clear[f];
f[r_, m_, 
  lambda_] := (r^2 (-1.000000000 + 
       2.000000000^(-1.000000000 Sqrt[
            1.000000000 + 0.3593734323 (1/m)^(2/3)]) (r/m)^
         Sqrt[1.000000000 + 
           0.3593734323 (1/m)^(2/3)]) (-1.000000000 (-1.000000000 + 
           Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)])^2 + 
       2.000000000^(-1.000000000 Sqrt[
            1.000000000 + 0.3593734323 (1/m)^(2/3)]) (1.000000000 + 
           Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)])^2 (r/m)^
         Sqrt[1.000000000 + 
           0.3593734323 (1/m)^(2/3)]))/(-1.000000000 + 
      Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)] + 
      2.000000000^(-1.000000000 Sqrt[
           1.000000000 + 0.3593734323 (1/m)^(2/3)]) (1.000000000 + 
         Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)]) (r/m)^
        Sqrt[1.000000000 + 0.3593734323 (1/m)^(2/3)])^2 + m^2 lambda^2

You can use RegionPlot3D

RegionPlot3D[
 f[r, m, lambda] > 0,
 {r, 0, 5},
 {m, 1, 5},
 {lambda, -1, 1},
 AxesLabel -> Automatic
 ]

Blockquote

Improve this answer
$\endgroup$
2
  • $\begingroup$ Yes, however now I must explicitly give a range for $r$. And the range for $r$ over which the roots occur (when they do) depends on $M,\lambda$ so then it becomes a little difficult to make inferences from this plot. $\endgroup$
    – newtothis
    Commented Oct 18, 2022 at 11:43
  • $\begingroup$ Rereading your question, what $λ$ and $M$ have no roots given the constraints of $0<λ<1$ and $M>1$ ? $\endgroup$
    – IntroductionToProbability
    Commented Oct 19, 2022 at 14:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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