Hi @kain88-de @jkoefinger ,

The method
rm = pydiffusion.rotation.RotationMatrix(u.atoms, ref=tetrahedron.atoms, verbose=True).run()
in the tutorial breaks when used with MDAnalysis v1.0.

The error comes from the
class RotationMatrix(AnalysisBase) in pydiffusion/rotation.py

The method _prepare() in RotationMatrix overwrites the self.frames definition of the parent class AnalysisBase, causing trouble for the latter in the run() method.

To fix this, I replaced all occurrences of self.frames in class RotationMatrix with self.Dframes in my local copy of rotation.py, corresponding to the following lines:

In method _prepare(self):
self.frames = [] --> self.Dframes = []

In method _single_frame(self):
self.frames.append(self._mobile.universe.trajectory.frame) --> self.Dframes.append(self._mobile.universe.trajectory.frame)

In method _conclude(self):
self.frames = np.asarray(self.frames) --> self.Dframes = np.asarray(self.Dframes)

This works, but there is perhaps a better way to address the issue.

Best,
Soeren

Hello,

I am trying to study the diffusion of a membrane embeded protein (KcsA specifically). I have been able to replicate the tutorial but I am getting some strange results with my system and I wonder if I am doing something wrong computationally or the theory doesn't work for these cases. I get these plots:

I understand why one of the main component doesn't converge since there is not a lot of vertical rotation of the protein. What I don't understand is why the other components don't converge to 0.25 as in the reset of examples in the paper and the tutorial.
Thank you very much for your help!
Best regards,
Sergio

The analysis class behavior has changed. So some code from the setup method needs to go into the prepare function.