The matrix of all first-order partial derivatives of a vector-valued function mapping between finite-dimensional Euclidean spaces; it represents the best linear approximation to the map at a point.

Principle

Local linearization: the Jacobian organizes how infinitesimal changes in input produce infinitesimal changes in output via the multivariable chain rule.

Demonstration

For F(x,y) = (u(x,y),v(x,y)), the Jacobian is [[∂u/∂x, ∂u/∂y],[∂v/∂x, ∂v/∂y]]; for polar coordinates (r,θ)↦(x=r cosθ,y=r sinθ) the Jacobian entries follow from those partial derivatives and its determinant equals r.

Misapplication

Using the Jacobian at a point where the mapping is not differentiable, or treating it as globally valid linearization for strongly nonlinear maps far from the expansion point.

Consequence

Enables local approximation of nonlinear maps, coordinate-change formulas in integrals via the Jacobian determinant, and linear stability analyses of fixed points.

Reversal

Instead of differentiating the forward map, one may consider the derivative of an inverse map; when invertible the inverse Jacobian equals the matrix inverse of the Jacobian.

Boundary

Requires differentiability (existence of partial derivatives) at the point; for infinite-dimensional mappings one needs an appropriate Fréchet or Gateaux derivative instead; singular points where the Jacobian is rank-deficient are excluded from invertibility claims.

Semantic Tension

Often confused with the Hessian (matrix of second derivatives) or a gradient (vector of first derivatives for scalar functions); the Jacobian is specifically the derivative of a vector-valued map.

Synthesis

The Jacobian matrix is the local linear map—assembled from first partial derivatives—that encodes how small input perturbations are transported to output perturbations and underpins coordinate changes and stability analysis.