Error estimates in EpsilonPlanar still broken?
I just did a test for the airwater system we have in the test file and the density errors look okay:
datapath = "../../../../tests/data/airwater"
u = mda.Universe(f'{datapath}/conf.gro', f'{datapath}/traj.trr')
dplan = maicos.DensityPlanar(u.select_atoms("name OW*"), dens='number').run()
plt.errorbar(dplan.results.z,
dplan.results.profile_mean.flatten(),
yerr=dplan.results.profile_err.flatten())
plt.ylabel("rho_N")
plt.xlabel("z")
But the epsilon error (here the parallel component) is too large as @schlaicha pointed out.
datapath = "../../../../tests/data/airwater"
u = mda.Universe(f'{datapath}/topol.tpr', f'{datapath}/traj.trr')
eplan = maicos.EpsilonPlanar(u.atoms).run()
plt.errorbar(eplan.results.z,
eplan.results.eps_par.flatten(),
yerr=eplan.results.deps_par.flatten())
plt.ylabel("eps_par")
plt.xlabel("z")
So this probably a problem solely of the epsilon module