Learn to Use ITensor

main / classes / spectrum C++v3 | C++v2


Note: this documentation page refers to code prior to version 2.0

A Spectrum object stores a reduced density matrix eigenvalue spectrum resulting from an SVD or density matrix decomposition. It is the return value of the svd and denmatDecomp methods. Spectrum objects can also be constructed from the diagonal singular-value tensor computed by the svd function.


Spectrum spec = svd(T,U,D,V);

Print(spec); //view the squares of the singular values of T

//Thinking of T as a wavefunction, compute its entanglement entropy S
Real S = 0;
for(int n = 1; n <= spec.numEigsKept(); ++n)
    S += -spec.eig(n)*log(spec.eig(n));
printfln("S = %.10f",S);

Back to Classes
Back to Main