Jacobian matrix

The Jacobian matrix $[\mathbf{J}f(\mathbf{a})]$ of a function $f\colon\mathbb{R}^{n}\rightarrow\mathbb{R}^{m}$ at the point $\mathbf{a}$ with respect to some choice of bases for $\mathbb{R}^{n}$ and $\mathbb{R}^{m}$ is the matrix of the linear map from $\mathbb{R}^{n}$ into $\mathbb{R}^{m}$ that generalizes the definition of the derivative of a function on $\mathbb{R}$. It can be defined as the matrix of the linear map $D$, such that

 $\lim_{\mathbf{h}\rightarrow\mathbf{0}}\frac{\|f(\mathbf{a+h})-f(\mathbf{a})-D(% \mathbf{h})\|}{\|\mathbf{h}\|}=0$

The linear map that satisfies the above limit is called the derivative of $f$ at $\mathbf{a}$. It is easy to show that the Jacobian matrix of a given differentiable function at $\mathbf{a}$ with respect to chosen bases is just the matrix of http://planetmath.org/node/841partial derivatives of the component functions of $f$ at $\mathbf{a}$:

 $[\mathbf{J}f(\mathbf{x})]=\left[\begin{array}[]{ccc}D_{1}f_{1}(\mathbf{x})&% \dots&D_{n}f_{1}(\mathbf{x})\\ \vdots&\ddots&\vdots\\ D_{1}f_{m}(\mathbf{x})&\dots&D_{n}f_{m}(\mathbf{x})\\ \end{array}\right]$

A more concise way of writing it is

 $[\mathbf{J}f(\mathbf{x})]=[\overrightarrow{D_{1}f}\;,\cdots\;,\overrightarrow{% D_{n}f}]=\left[\begin{array}[]{c}\nabla f_{1}\\ \vdots\\ \nabla f_{m}\end{array}\right]$

where $\overrightarrow{D_{n}\mathbf{f}}$ is the partial derivative with respect to the $n$โth variable and $\nabla f_{m}$ is the gradient of the $m$โth component of $\mathbf{f}$.

Note that the Jacobian matrix represents the matrix of the derivative $D$ of $f$ at $\mathbf{x}$ iff $f$ is differentiable at $\mathbf{x}$. Also, if $f$ is differentiable at $\mathbf{x}$, then the directional derivative in the direction $\vec{v}$ is $D(\vec{v})=[\mathbf{J}f(\mathbf{x})]\vec{v}$.

Given local coordinates for some real submanifold $M$ of $\mathbb{R}^{n}$, it is easy to show that the effect of a change of coordinates on volume forms is a local scaling of the volume form by the determinant of the Jacobian matrix of the derivative of the backwards change of coordinates, which is called the inverse Jacobian. The determinant of the inverse Jacobian is thus commonly seen in integration over a change of coordinates.

 $\int_{\Omega_{x}}f(x)d\Omega_{x}=\int_{\Omega_{\xi}}f(\xi)|J^{-1}|d\Omega_{\xi}$

where $x\in\Omega_{x}\subset M$, $\xi\in\Omega_{\xi}\subset M$, and $\mathbf{J}=\nabla\xi(x)$ is the derivative of the function $\xi(x)$ which maps points in $\Omega_{x}$ to points in $\Omega_{\xi}$.

 Title Jacobian matrix Canonical name JacobianMatrix Date of creation 2013-03-22 11:58:33 Last modified on 2013-03-22 11:58:33 Owner PhysBrain (974) Last modified by PhysBrain (974) Numerical id 18 Author PhysBrain (974) Entry type Definition Classification msc 26B10 Synonym derivative of a multivariable function Synonym derivative of a vector-valued function Synonym Jacobi matrix Related topic PartialDerivative Related topic Derivative2 Related topic DerivativeNotation Related topic Gradient Related topic ChainRuleSeveralVariables Related topic DirectionalDerivative Related topic JordanBanachAndJordanLieAlgebras Defines Jacobian