For example, I have: $\lim_{(x,y)\to(0,0)}\frac{e^{x^2+y^2}\ln(2(x^2+y^2)+1)-\frac{3}{2}(x^2+y^2)-1}{x^2+y^2}$
After a change to polar coordinates: $\lim_{r\to0}\frac{e^{r^2}\ln(2(r^2)+1)-\frac{3}{2}(r^2)-1}{r^2}$
I think that since the angle disappears the function has some sort of radial symmetry, so it doesn't matter how do I approach to $(0,0)$. After that I solved the limit like I would solve a single variable limit using L'Hopital's and the limit was equal to $\frac{3}{2}$ which is correct. Am I correct in assume that every time the angle disappears I can evaluate the limit as if it were one variable? Or what happened there? (Sorry if I have some spelling mistakes, I'm not english native speaker).
Yes, every time the variables show up they show up in the form of $r^2$ as your transformation shows, so in polar the angle is indeed not needed and you get a one variable limit