Analyzing Stationary Points of Multivariable Function

  • Context: MHB 
  • Thread starter Thread starter Pulty
  • Start date Start date
  • Tags Tags
    Points
Click For Summary

Discussion Overview

The discussion focuses on analyzing stationary points of the multivariable function $$f(x,y) = \frac{1}{2}{x}^{2}{e}^{y}-\frac{1}{3}{x}^{3}-y{e}^{3y}$$. Participants explore the identification of critical points, the application of the second partials test, and the classification of these points as relative extrema or saddle points.

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant presents the function and its partial derivatives, seeking to identify stationary points.
  • Another participant confirms the need to equate the first partial derivatives to zero and identifies two critical points: $$(0, -\frac{1}{3})$$ and $$(e^{-\frac{1}{6}}, -\frac{1}{6})$$.
  • Subsequent posts compute second partial derivatives and evaluate the determinant for both critical points, with one point classified as a saddle point and the other as a relative maximum.
  • Participants discuss the method for finding the critical point $$(0, -\frac{1}{3})$$, with one participant explaining their approach to derive it from the equations.

Areas of Agreement / Disagreement

There is agreement on the identification of the critical points and the application of the second partials test. However, the classification of the points as extrema or saddle points is based on individual calculations, and no consensus is reached on the implications of these classifications beyond the presented evaluations.

Contextual Notes

Participants rely on specific calculations of second partial derivatives and determinants, which may depend on the accuracy of earlier steps. There are no explicit corrections or resolutions to potential uncertainties in the calculations presented.

Who May Find This Useful

Readers interested in multivariable calculus, particularly in the analysis of stationary points and the application of the second derivative test, may find this discussion relevant.

Pulty
Messages
12
Reaction score
0
$$f(x,y) = \frac{1}{2}{x}^{2}{e}^{y}-\frac{1}{3}{x}^{3}-y{e}^{3y}$$

To start off I found the partial derivatives of

$$x: {e}^{y}x - {x}^{2}$$

$$y: \frac{{e}^{y}{x}^{2}}{2}-{e}^{3y}-3{e}^{3y}y$$

Then solved simultaneously for each equation equal to 0.

$$y = \ln(x) = -\frac{1}{6}$$

$$x = {e}^{-\frac{1}{6}}$$

(Is there another stationary point?)

I'm unsure of where to go from here, or if I'm even doing the right thing to begin with.

Thank you
 
Last edited by a moderator:
Physics news on Phys.org
You are correct in that you first want to identify all critical points, by equating the first partials to zero:

$$f_x(x,y)=xe^y-x^2=x\left(e^y-x\right)=0$$

$$f_y(x,y)=\frac{1}{2}x^2e^y-\left(3ye^{3y}+e^{3y}\right)=\frac{1}{2}e^{3y}\left(x^2e^{-2y}-6y-2\right)=0$$

And solving them simultaneously.

From these we obtain:

$$(x,y)=\left(0,-\frac{1}{3}\right),\,\left(e^{-\frac{1}{6}},-\frac{1}{6}\right)$$

So, you had one point, you just missed when $x=0$ from $f_x=0$. :)

Now, armed with these two points, you want to apply the second partials test for relative extrema. Essentially, you want to compute:

$$D(x,y)=f_{xx}(x,y)f_{yy}(x,y)-\left[f_{xy}(x,y)\right]^2$$

What do you get?
 
fxx(x,y) = e^y - 2x

fyy(x,y) = -9e^(3y)y - 6e^(3y) + (e^(y)+x^(2))/2

fyx(x,y) = e^(y)x

At point (0,-1/3):

fxx(x,y)fyy(x,y) - fxy(x,y)^2 = -0.79079 (Saddle point)

At point (e^1/6, -1/6)

fxx(x,y)fyy(x,y) - fxy(x,y)^2 = 3.08049 (Maximum point)

How does this look?

Also, I am confused as to how you find the point (0,-1/3)

Thanks
 
Computing the second partials, we find:

$$f_{xx}(x,y)=e^y-2x$$

$$f_{yy}(x,y)=\frac{1}{2}x^2e^y-\left(\left(9ye^{3y}+3e^{3y}\right)+3e^{3y}\right)=\frac{1}{2}x^2e^y-3e^{3y}\left(3y+2\right)=\frac{1}{2}e^{3y}\left(x^2e^{-2y}-6(3y+2)\right)$$

$$f_{xy}(x,y)=xe^y$$

And so for the first critical point, we have:

$$f_{xx}\left(0,-\frac{1}{3}\right)=e^{-\frac{1}{3}}$$

$$f_{yy}=-\frac{3}{e}$$

$$f_{xy}\left(0,-\frac{1}{3}\right)=0$$

$$D\left(0,-\frac{1}{3}\right)=e^{-\frac{1}{3}}\left(-\frac{3}{e}\right)-0^2=-3e^{-\frac{4}{3}}<0$$

This point is not an extremum.

And for the second critical point, we have:

$$f_{xx}\left(e^{-\frac{1}{6}},-\frac{1}{6}\right)=-e^{-\frac{1}{6}}$$

$$f_{yy}\left(e^{-\frac{1}{6}},-\frac{1}{6}\right)=\frac{1}{2}e^{-\frac{1}{2}}\left(e^{-\frac{2}{3}}-9\right)$$

$$f_{xy}\left(e^{-\frac{1}{6}},-\frac{1}{6}\right)=e^{-\frac{1}{3}}$$

$$D\left(e^{-\frac{1}{6}},-\frac{1}{6}\right)=\left(-e^{-\frac{1}{6}}\right)\left(\frac{1}{2}e^{-\frac{1}{2}}\left(e^{-\frac{2}{3}}-9\right)\right)-\left(e^{-\frac{1}{3}}\right)^2>0$$

This point is a relative maximum.

So, I agree with your conclusions regarding the critical points. :D

How I found the other critical point was from:

$$f_x(0,y)=0$$

And so we then set:

$$f_y(0,y)=-e^{3y}(3y+1)=0\implies y=-\frac{1}{3}$$

And so we find the point:

$$(x,y)=\left(0,-\frac{1}{3}\right)$$

simultaneously satisfies:

$$f_x(x,y)=0$$

$$f_y(x,y)=0$$

as required.
 
Thanks Mark!

I have a similar question I will ask in a new thread.
 

Similar threads

  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 9 ·
Replies
9
Views
1K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 16 ·
Replies
16
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 27 ·
Replies
27
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K