Is There a Homogenous Function of Degree n < 0?

  • Context: Graduate 
  • Thread starter Thread starter psholtz
  • Start date Start date
  • Tags Tags
    Functions
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
2 replies · 3K views
psholtz
Messages
133
Reaction score
0
Is there such a thing as a homogenous function of degree n < 0?

Considering functions of two variables, the expression:

[tex]f(x,y) = \frac{y}{x}[/tex]

is homogeneous in degree 0, since:

[tex]f(tx,ty) = \frac{ty}{tx} = \frac{y}{x} = f(x,y) = t^0 \cdot f(x,y)[/tex]

and the expression:

[tex]f(x,y) = x[/tex]

is homogenous in degree 1 since:

[tex]f(tx,ty) = tx = t^1 \cdot f(x,y)[/tex]

and the expression:

[tex]f(x,y) = x^3y^2[/tex]

is homogeneous in degree 5, since:

[tex]f(tx,ty) = t^5 \cdot f(x,y)[/tex]

I suppose that in the same way I could construct a function something like:

[tex]f(x,y) = \frac{y}{x^2}[/tex]

So that:

[tex]f(tx,ty) = t^{-1} \cdot f(x,y)[/tex]

or in other words, "homogenous" in degree n = -1.

Does this ever really come up much?
 
Physics news on Phys.org
I think the answer is "yes"..

For instance, the Euler theorem on homogeneous functions states, in relevant part, that if you have a function in two variables, x and y, which is homogeneous in degree n, then:

[tex]n f(x,y) = x \frac{\partial f}{\partial x} + y \frac{\partial f}{\partial y}[/tex]

So taking the example I gave above, where you have:

[tex]f(x,y) = \frac{x}{y^2}[/tex]

which, ostensibly, is homogeneous in degree -1. Plugging this equation in the Euler theorem above, you would have:

[tex]\frac{\partial f}{\partial x} = \frac{1}{y^2}[/tex]

[tex]x\frac{\partial f}{\partial x} = \frac{x}{y^2}[/tex]

and that:

[tex]\frac{\partial f}{\partial y} = -2\frac{x}{y^3}[/tex]

[tex]y\frac{\partial f}{\partial y } = -2\frac{x}{y^2}[/tex]

so that:

[tex]x\frac{\partial f}{\partial x} + y\frac{\partial f}{\partial y} = -\frac{x}{y^2} = -f(x,y)[/tex]

which is what we would expect of a function which is homogenous in degree -1..
 
Yes.

Consider the following:

[tex] U = \sum_{1 \leq i < j \leq n}^{n} \frac{m_i m_j}{||q_i - q_j||}[/tex]

It usually appears in the n-body problem in the form of...

[tex] m_i \ddot{x_i} = \frac{dU}{dq_i}[/tex]

The fact that U is homogeneous of degree -1 is key to prove that there are no equilibrium solutions for the n-body problem.