I'm currently working on a problem for homework in my Banach algebras course and I've run into a bit of an issue with terminology. Let $\mathcal{A}$ be a Banach algebra, then $\mathcal{A}^{-1}_0$ denotes the subgroup of invertible elements that can be written as a finite product of exponentials. That is to say that if $a\in\mathcal{A}^{-1}_0$, then $a=e^{b_1}\cdots e^{b_n}$, where $b_1,\ldots,b_n\in\mathcal{A}$.

My question is for clarification as to what the following means (I do not want help with the actual proof, just clarification): Show that $\mathcal{A}^{-1}_0$ consists of all invertible elements which can be connected to the identity by a continuous path of invertible elements.

I'm aware that such an idea exists in Lie theory but I've never studied that in any meaningful way so paths of elements isn't a well-defined concept for me.


Well, as a Banach algebra, $\mathcal A$ has a norm, so naturally a topology, hence 'continuous path' does make sense here.

One direction of the statement is easy: if $a=e^{b_1}e^{b_2}\dots e^{b_n}$, then the continuous function $[0,1]\to\mathcal A,\ \,t\mapsto e^{tb_1}e^{tb_2}\dots e^{tb_n}\ \,$ connects $1$ to $a$.

  • $\begingroup$ Oh I see what continuous path means now. I wasn't really sure what that term meant. Thanks. $\endgroup$ – Cameron Williams Jan 26 '14 at 23:58
  • $\begingroup$ Though, I cannot see yet the other direction. Can you share with me if you find it? $\endgroup$ – Berci Jan 27 '14 at 0:01
  • $\begingroup$ I will if/when I figure it out! $\endgroup$ – Cameron Williams Jan 27 '14 at 0:02
  • $\begingroup$ I guess it would require some differentiation somehow.. Probably the topology induced by the norm is nice enough that every continuous path can be transformed arbitrarily close to a differentiable path. $\endgroup$ – Berci Jan 27 '14 at 0:06

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.