Let $G$ be a topological group, and $X$ some function space; preferably a Sobolev space $X=W^{1,p}(\Omega)$, where $\Omega \subset \mathbb{R}^n$ is some invariant subset ($g\Omega \subset \Omega$) or the whole space.
Let $G$ be represented as linear, continuous operators $\pi(g)$ on $X$ in the form $$ \pi(g)f(x)=f(gx) $$
Does the set of invariant functions that satisfy $$ f(gx)=f(x) $$ form a Banach (sub)manifold of $X$?
Do we need additional assumptions on the group (e.g.compact groups)? As an example consider $G=O(n)$ the orthogonal group. Does the set of radial symmetric functions form a Banach (sub)manifold of $X$?
Feel free to modify and add assumptions if necessary as I am not too familiar with this field. I am thankful for any reference, hint or remark!


In most cases they form a closed subspace: it is subspace, because if $f$ and $g$ are invariant, then so is $f+g$ or $\lambda f$. Whenever it is closed or not depends on norm, but in case of $\sup$-norm or $L^2$ norm it is.

The subspace of such functions can be considered as functions on space of orbits $\Omega/G\Omega$, where measure induced from $\Omega$.

  • $\begingroup$ Thank you for your answer! One more question: Do you have any recommendation in terms of literature? I am also interest in equivariant maps. $\endgroup$ – F. Conrad May 14 '19 at 23:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.