Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm reading this book and I'm stuck on page 223. Since some pages are missing, let me describe the problem formulation.

Preliminaries: Let $f:\mathbb R^n\to\mathbb R^n$ be $C^1$ vector field and consider an ODE $$ \dot x = f(x)\quad (1) $$ with a correspondent flow $\phi(t,x)$ such that $$ \frac{\partial \phi}{\partial t}(t,x) = f(\phi(t,x)) $$ and $\phi(0,x) = x$. Furthermore, let $$ Dg = \left(\frac{\partial g_i}{\partial x_j}\right)_{i,j=1}^n $$ denote the Jacobian matrix of $g:\mathbb R^n\to\mathbb R^n$.

Question: the matrix $H(t,x)$ is given by $H(t,x) = D\phi(t,x)$ where the Jacobian is w.r.t. $x$ coordinates only. It's written on the top of p. 223 that $$ \frac{\partial H}{\partial t}(t,x) = Df(\phi(t,x))H(t,x) $$ and I cannot understand why do we have $H(t,x)$ multiplier.

My thoughts are the following: $$\begin{split} \frac{\partial H_{ij}}{\partial t}(t,x) &= \frac{\partial}{\partial t}\frac{\partial \phi_{i}}{\partial x_j}(t,x) = \frac{\partial }{\partial x_j}\frac{\partial \phi_i}{\partial t} = \frac{\partial}{\partial x_j}f_i(\phi(t,x))\\&=\sum_{k=1}^n\frac{\partial f_i}{\partial \phi_j}(\phi(t,x))\frac{\partial \phi_k}{\partial x_j}(t,x) \end{split}$$

Are my calculations correct?

share|cite|improve this question
up vote 2 down vote accepted

Yes, they are correct.

The fact that $H$ comes up again on the RHS means that you have successfully linearised your initially nonlinear differential equation. So by "pretending" that $\phi$ and $H$ are independent quantities (when in reality, they are highly dependent on each other), you can use techniques of linear differential equations to get a lot of information on $H$, provided some small bits of information is assumed to be true for $\phi$.

That you can "assume a little and gain a lot more" is what then powers the analysis of the actual solution $\phi$.

share|cite|improve this answer
Thanks, I'm familiar with Floquet's Theory - but apparently hasn't calculated derivatives for a long time. Namely, I realized the place when the chain rule appears only while typing the question - before I was missing the point that we should take derivative of $f$ w.r.t. $\phi_k$ rather that w.r.t. $x_k$. This fact always bothers me in partial derivatives of compositions – Ilya Dec 20 '11 at 15:12

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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