How to compare the magnitude (the absolute value, ignoring the sign) of two multi-parameter functions f_1 and f_2?

Note that all my parameters can only take stricly positive values. Specifically, I want to know the ranges of parameter values for which f_1 > f_2, f_1 < f_2, and f_1 = f_2. Is solve(f_1 > f_2) etc. the best way to do this? If so, how to specify solve() so that it incorporates all three (>,<,=) comparisons and do not ignore the strict positivity assumption? Is there any other command other than solve()?   

Example with only two parameters sigma__v and sigma__d:

Download parametric_comparison.mw

Note that f_1 depends only on sigma__d but f_2 depends on both sigma__v and sigma__d. Of course I can directly plot the two curves in 3d in this simple case, but I am looking for a systematic way to do parametric comparisons so that I don't have to "eyeball" the threshold values (if any exist). Most importantly, I can't plot at all once my functions depend on more than two parameters.

 Thanks.

EDIT: perhaps "variables" is a better word than "parameters" here.

I want to approximate a positive function that is decreasing in Gamma, say f(Gamma), that is very complicated yet very smooth. I need this in order to obtain a tractable and compact version of its derivative, which enters in the partial derivative of another (very simple) function.

Along the way, three related questions emerge: Derivatives_and_Approximations.mw

Thanks a lot!

Hi,

How can two specular equations generate two nonspecular (i.e., very different in length and form) solutions? I attach my script with two questions at the bottom: 

specular_equations_nonspecular_solutions.mw

Thank you.

I am re-posting. I am not sure why my question was deleted. Please advise on how to amend my post so that it's not considered spam.

I simply wonder how parameters can "disappear" in a solution. In particular, in my example below the parameter gamma correctly appears in 3 out of 4 solution. However, the solution in which 'gamma' does not appear is also the solution I am most interested in, given its manageable size. Why gamma is "lost" for this solution? 281223_gamma_disappear.mw

I am solving 3 nonlinear equations for 3 variables: lambda_1, lambda_2, and lambda_3. I would expect these lambdas to be real and positive.

Instead of solving my original equations, which are convoluted and not in polynomial form, I try to solve for their numerators first (since their numerators are polynomials). Broadly speaking, such solutions should also solve the original non-polynomial system. More specifically, the solutions thus obtained may be a nontrivial superset of the solutions of the original system. They need to be verified, which should be a much much easier process than obtaining that superset. In the case at hand, my original system is rational functions, and thus the only thing that really needs to be verified is that the solutions do not make any of the original denominators zero.

1st question: How to actually implement such verification? In other words, how to verify that the polynomial solution that I obtain also solves the original non-polynomial system?

2nd question: As you can see from my attached script, I obtain one polynomial solution. How to analyze it? What can I say about its roots? In case there are an infinite number of roots, how can I pin down a closed-form, real, and positive expression of lambda_1, lambda_2, lambda_3 in terms of the four parameters gamma, p, sigma_e and sigma_v?*

*Please note that in SolveTools:-PolynomialSystem I set backsubstitute=false to favour compactness and computational efficiency (which means that I need to do the backsubstitution myself now - how to do it?).

**Perhaps is useful to know that gamma, sigma_e and sigma_v are all real and positive and that p is a real, positive number between 0 and 1 (it represents a probability).

SCRIPT: 141123_Problem_NoCorrelation.mw

Thanks a lot!