Is there a particular reason why? Not sure if I’m missing something obvious. Tagging @rkingsbury since he was asking.
I don’t know any particular reason why - maybe someone figured that residual forces are only interesting if you are doing structure optimization? But I can certainly see the use to having these parsed for static runs as well.
Yeah, we were doing some static convergence testing and wanted to see how consistently forces were being evaluated. I can submit a PR. Thanks!
@rkingsbury Turns out forces are parsed for static calculations, they’re in output.forces, it’s only the analysis.max_force key that isn’t present but you can still find the max force manually via max(np.linalg.norm(forces, axis=1)).