DistanceMatrix accept AtomGroup

[lilyminium]

Is your feature request related to a problem?

Is there any reason DistanceMatrix cannot accept an AtomGroup inplace of a Universe? This would help with atomgroups created through non-selection-string ways, which you then need to re-create with atom indices. e.g.
https://groups.google.com/g/mdnalysis-discussion/c/ad4HW_6w2MM/m/-h3bRDv0DgAJ?utm_medium=email&utm_source=footer

Describe the solution you'd like

Here:

mdanalysis/package/MDAnalysis/analysis/diffusionmap.py

Lines 251 to 253 in 5bf6359

	super(DistanceMatrix, self).__init__(universe.trajectory, **kwargs)
	self.atoms = universe.select_atoms(select)

- universe.trajectory
+ universe.universe.trajectory

Describe alternatives you've considered

Additional context

[richardjgowers]

AtomGroup is probably preferable everywhere to Universe as an argument imho. Universe means a lot of things, AG is clearer.

…

On Mon, Dec 27, 2021 at 13:26, Lily Wang ***@***.***> wrote: Is your feature request related to a problem? Is there any reason DistanceMatrix cannot accept an AtomGroup inplace of a Universe? This would help with atomgroups created through non-selection-string ways, which you then need to re-create with atom indices. e.g. https://groups.google.com/g/mdnalysis-discussion/c/ad4HW_6w2MM/m/-h3bRDv0DgAJ?utm_medium=email&utm_source=footer Describe the solution you'd like Here: https://github.com/MDAnalysis/mdanalysis/blob/5bf6359a821c085a1039ca414ce51ca2820722fb/package/MDAnalysis/analysis/diffusionmap.py#L251-L253 - universe.trajectory+ universe.universe.trajectory Describe alternatives you've considered Additional context — Reply to this email directly, view it on GitHub <#3486>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACGSGB574NRBZKLKON7K2PDUTBQPXANCNFSM5K2KVOZA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[ceaserbutdumb]

Hey, I am a beginner here so could you guide me to what I can do to solve this issue?

[IAlibay]

@ceaserbutdumb in the first instance the minimal change required here would be to just change the documentation and __init__ code to allow for an AtomGroup to be used instead of universe. For the most part it should work as-is, but you'll need to add some test to make sure that it does what it should do.

It's hard for us to give specific guidance about these things, if you can open a pull request with an attempt at achieving this task, then we can try to help push the pull request along towards a merge.

[ceaserbutdumb]

@IAlibay I will try to work on it and thank u for the help

[IAlibay]

@lilyminium shouldn't here we have Universe.trajectory instead of universe.trajectory

@ceaserbutdumb may I ask you that you do not write comments and then delete them please? Especially when you ping an individual directly this makes it difficult for us to work out what happened (i.e. we get an email notification but there's no comment when we attempt to reply on github).

[ceaserbutdumb]

@lilyminium shouldn't here we have Universe.trajectory instead of universe.trajectory

@ceaserbutdumb may I ask you that you do not write comments and then delete them please? Especially when you ping an individual directly this makes it difficult for us to work out what happened (i.e. we get an email notification but there's no comment when we attempt to reply on github).

I am sorry

[abhik-123]

@IAlibay I am new to open source and i wish to contribute in this project. Could you please suggest how should I begin?
I am good at coding in python

[IAlibay]

@IAlibay I am new to open source and i wish to contribute in this project. Could you please suggest how should I begin? I am good at coding in python

In the first instance I would recommend our user guide as your first point of call for getting to know MDAnalysis: https://userguide.mdanalysis.org/stable/index.html. The user guide includes a section on how to contribute to the project.

I would ask that questions about how to get started / how to contribute / etc... be asked on the MDAnalysis mailing lists or our discord server rather than issues. Please see here for more details: https://www.mdanalysis.org/#participating.

If you have questions specific to this particular issues, then please do ask them here.

[AnirG]

Hey @richardjgowers @lilyminium!

AtomGroup is probably preferable everywhere to Universe as an argument
In the first instance the minimal change required here would be to just change the documentation and init code to allow for an AtomGroup to be used instead of universe.

I'm not completely sure how this can still work. Without the universe passed on as a parameter, how can trajectories be looped over to update our AtomGroup which is needed to calculate pairwise metrics?
Apologies if it's a naive question, I'm kinda new to this stuff.

[richardjgowers]

@AnirG AtomGroup.universe.trajectory

[AnirG]

@richardjgowers Ah cool, I see! I can possibly try to work on this one. I will also try to code up some test cases. Should the changes and test cases be raised in a single PR?