3

Show that the limit as $(x,y)\to (0,0)$ does not exist for the function $$ f(x,t)=\begin{cases}\frac{x^3+y^3}{x-y},\quad & x\ne y \\ 0,\quad & x=y\end{cases} $$

My solution is below, but it appears to be inconsistent with the problem. What is wrong?

enter image description here

enter image description here

godonichia
  • 2,259
  • 1
  • 25
  • 47

1 Answers1

4

Your conversion to polar coordinates does not prove the limit exists, because in doing so, you are assuming that as $r \to 0^+$, that $\theta$ is a fixed constant with respect to the change in $r$, thus you are only considering a subset of all possible paths tending toward $(0,0)$--specifically, the subset of all straight-line paths to the origin.

See this post for a similar problem, and more information: Limit is found using polar coordinates but it is not supposed to exist.

So, the prove that the limit does not exist, it suffices to discover two distinct paths to $(0,0)$, each suitably parametrized, such that the limit along these paths are unequal to each other.

heropup
  • 143,828