If $||\cdot||$ is a matrix norm and $A$ is a matrix, is there any relation between the two quantities $||A||$ and $||e^A - I||$, where $I$ is the identity matrix?

I am tempted to say the former is always smaller than or equal to the latter, analogous to $x \leq e^x - 1$ for all $x \in \mathbb{R}$, but my numerical experiment suggests that it is the other way: $||A|| \geq ||e^A - I||$. Can some one please provide some idea?

  • $\begingroup$ There is no such relationship for the Euclidean norm. Consider $A=I$ and $A=-I$ which give $\le$ and $\ge$ respectively. $\endgroup$ – Peter Foreman Jan 15 at 22:40

If the norm is submultiplicative (i.e. $\ \|AB\,\|\le \|A\,\| \|B\,\|\ $ for all $\ A,B\ $), then you do get \begin{align} \left\|e^A-I\right\|&= \left\|\sum_{n=1}^\infty\frac{A^n}{n!} \right\|\\ &\le \sum_{n=1}^\infty\frac{\|A\|^n}{n!}\\ &=e^{\|A\|}-1\ , \end{align} but since $\ e^x\ $ is not bounded above by any polynomial in $\ x\ $, neither is $\ e^{\|A\|}-1\ $ bounded above by any polynomial in $\ \|A\|\ $.

| cite | improve this answer | |

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.