Suppose $A$ is a $n \times n$ constant matrix. How can I prove $\det(e^A) = e^{\displaystyle \sum_{\lambda_i\in\sigma(A)} \lambda_i}$,

where $\sigma(A)$ is the multiset of eigenvalues of $A$?

The following matlab code shows this is true in $n = 3$ case:

A = rand(3)
detA = exp(sum(eig(A)))
detmA = det(expm(A))
  • 4
    $\begingroup$ prove it for diagonal matrices and use denseness of diagonalizable matrices. notice that the trace of $A$ is the sum of the eigenvalues $\endgroup$ – cats Mar 1 '13 at 20:34

If you're afraid of the density of diagonalizable matrices, simply triangularize $A$. You get $$A=P^{-1}UP,$$ with $U$ upper triangular and the eigenvalues $\{\lambda_j\}$ of $A$ on the diagonal.

Then $$ \mbox{det}\;e^A=\mbox{det}(P^{-1}e^UP)=\mbox{det}\;e^U. $$

Now observe that $e^U$ is upper triangular with $\{e^{\lambda_j}\}$ on the diagonal.

So $$ \mbox{det} \;e^A=\mbox{det} \;e^U=e^{\lambda_1}\cdots e^{\lambda_n}=e^{\lambda_1+\ldots+\lambda_n}. $$

  • 5
    $\begingroup$ Anyone trying to understand this solution: The main idea is that if the eigenvalues of $A$ are $\lambda_i$, then the eigenvalues of $e^A$ are $e^{\lambda_i}$, which you can see by using the series definition of $e^A$ and multiplying nonzero eigenvector $v_i$ as in $e^A v_i = \lambda_i v_i$. Distribute the $v_i$ through the series. Finally, the determinant of a matrix is the product of the eigenvalues, and the trace of a matrix is the sum of the eigenvalues. This explains the second to last equality in the proof above. $\endgroup$ – wkschwartz Oct 25 '15 at 23:28

You can reduce problem for diagonal matrix using the fact that every matrix can be approximated with any given precision and both functions: $e^X$ and $\det X$ are continuous. The problem with diagonal matrix is obvious.

P.S. It seems that lyj was faster than me.


For an analytic method, using differential equations:

Let $f(t)= \det(e^{tA})$. Then $f'(t)=D \det(e^{tA}) \cdot Ae^{tA}=\text{tr} \left(^t \text{com}(e^{tA})Ae^{tA} \right)$. But $A$ and $e^{tA}$ commute, and $^t\text{com}(e^{tA})e^{tA}=\det(e^{tA}) \operatorname{I}_n$. Therefore, $f'(t)=\text{tr}(A)f(t)$ and $f(0)=1$, hence $f(t)=e^{\text{tr}(A)t}$. For $t=1$, $\det(e^{A})= e^{\text{tr}(A)}$.

  • $\begingroup$ What kind of commutator is $^t\text{com}(e^{tA})$? Never seen this notation. $\endgroup$ – Your Majesty Jan 29 '14 at 10:13
  • 1
    $\begingroup$ $\mathrm{com}(A)$ is the matrix of cofactors associated to $A$. Possibly, it is a French notation, where $\mathrm{com}$ stands for comatrice. $\endgroup$ – Seirios Jan 29 '14 at 11:56
  • 2
    $\begingroup$ Hello from 2016! Could you please give a reference to this proof? $\endgroup$ – Rubi Shnol May 17 '16 at 6:35

You just use the Jordan normal form, $A$ is your matrix, $D$ is the Jordan normal form of $A$ and $S$ is the transformation matrix, for the second equal just remember the definition of matrix exponential, and that $$(SAS^{-1})^n = S A S^{-1}S A S^{-1} S A \dots S A S^{-1}=S A^n S^{-1}$$ \begin{align*} \det(\exp(A))&=\det(\exp(S D S^{-1}))\\ &=\det(S \exp(D) S^{-1})\\ &=\det(S) \det(\exp(D)) \det (S^{-1})\\ &=\det(\exp (D))\\ &=\prod_{i=1}^n e^{d_{ii}}\\ &=e^{\sum_{i=1}^n{d_{ii}}}\\ &=e^{\text{tr}D} \end{align*} As the trace is invariant this works.


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.