A Question about Normal Mode Sampling

[MasterLucas]

Dear ANI developers,

I have a question about the normal mode sampling you developed in the first ANI paper. First, is the code in the GitHub repository, I couldn't find it. Second, in the paper, what do you mean in this snippet "Next, a new conformation of the molecule is generated by displacing the energy minimized coordinates by Q^R, the superposition of all q_i^R." Should I sum the displacement of all the normal modes and then add it to the relaxed structure, or should I add the displacement of each normal mode to the relaxed structure, so that, in this case, if my molecule has 3 normal modes, the output will be three new rattled geometries, whereas in the former it would have only one rattled geometry?

Yours faithfully,