I need to apply a differential operator (nabla) on a matrix. Problem is, that I don't know how to calculate that. Do I treat nabla as a column vector and simply multiply vector with the matrix? Or is nabla to be handled as jacobi matrix?

  • $\begingroup$ I've never seen the standard del operator $\nabla$ applied to a matrix. What is the context of the application? $\endgroup$
    – Emily
    Commented Aug 23, 2012 at 18:05
  • $\begingroup$ There isn't really any difference between the $\nabla$ and Jacobi-matrix (up to transposition), it only depends on where your function, defined on matrices, maps to. $\endgroup$
    – Cocopuffs
    Commented Aug 23, 2012 at 18:11
  • $\begingroup$ It depends on what you mean by applying a differential operator to a matrix. What kind of entries does the matrix have? What do you want this differential operator to do to it? $\endgroup$ Commented Aug 23, 2012 at 18:12
  • $\begingroup$ Context is the nonlinear sigma model in quantum field theory, where I want to use a matrix function $\phi$ instead of a vector function. The matrix is a 3x3 one with simple scalar functions in it. $\endgroup$
    – Tornado
    Commented Aug 23, 2012 at 18:13
  • 1
    $\begingroup$ Treat it the same as a vector function, just with the entries in a different arrangement. $\endgroup$
    – Cocopuffs
    Commented Aug 23, 2012 at 18:17

1 Answer 1


The differential operators can be defined in a matrix of differential operators, so that, for instance, it could be applied to a matrix containing n vectors, one in each column.

$$ \left[\begin{matrix} \frac{\partial}{\partial{x}} & \frac{\partial}{\partial{y}} & \frac{\partial}{\partial{z}}\\ \end{matrix}\right] \left[\begin{matrix} x_1 & x_2 & x_3 & & x_n \\ y_1 & y_2 & y_3 & \dots & y_n \\ z_1 & z_2 & z_3 & & z_n \\ \end{matrix}\right] $$

To see how to implement this kind of operator in the Python mudule Sympy, you can refer to here. To see more discussion about matrix of differential operator, see here.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .