In the attachment I pose a problem of representing and computing with a matrix-valued function of a vector. Before I go to the effort of learning Mathematica, I would like to know that it can do these kinds of computations and yield results in the usual notation or something close to it. I would appreciate any guidance, including being told that my request is impossible. Thanks.
Mathematica matrix problem.pdf (38.2 KB)
It can be done:
m[\[Sigma]_] := With[{coeff = 1/Sqrt[1 - \[Sigma].\[Sigma]]},
ArrayFlatten[{{coeff, {coeff*\[Sigma]}},
{coeff*Transpose[{\[Sigma]}],
IdentityMatrix[Length[\[Sigma]]] +
(coeff - 1) 1/\[Sigma].\[Sigma]*
KroneckerProduct[\[Sigma], \[Sigma]]}}]];
m[{a, b, c}]
You have to get used to an unusual way of thinking about row and column vectors.
To Gianluca: Thank you for the Mathematica code. I will definitely find a way to try this.