Jennifer A Clark

It seems that when chain.py uses activate_reader.ts, it is using a separate Timestep (different id() ) with the Timestep instance of the universe. Thus, when wrap.wrap() called the AtomGroup.wrap function,...

Hi @richardjgowers, the issue seems to specifically be with the `wrap` transformation. The chain reader will (1) activate the new `ts`, (2) transform the `ts` trajectory, and then (3) associate...

This issue can be resolved with the following two lines of code: `MDAnalysis.transformations.wrap()` ``` def _transform(self, ts): + if id(ts) != id(self.ag.universe.trajectory.ts): + self.ag.universe.trajectory.ts = ts self.ag.wrap(compound=self.compound) return ts ```

@orbeckst it was brought up in the review of my Pull Request that this could also be an issue in the unwrap function. Indeed this is the case. If I...

Using the same example given in this issue, I looked at `fit_translation` and `fit_rot_trans` and I do think it belongs in this Pull Request. To reproduce what I'm describing, run...

Hey sure! I can get to it next week. > On May 12, 2022, at 2:34 AM, Oliver Beckstein ***@***.***> wrote: > >  > @jaclark5 if you've already identified...

This issues were resolved in this fork #[https://github.com/jaclark5/mdanalysis/tree/xyz_writer](url)

I don't see the harm in adding a keyword argument as most users won't even know it's there. There's a package called Zeno that our group uses that takes XYZ...

It seems I caused more problems by bringing up my application with variable atoms. Regardless, I didn't think exposing an inherent option would be an issue.

Fair enough, I guess I wouldn't need to open and close it multiple times if I could write a different number of atoms along the trajectory, but it seems that...