add an observable for pairwise distance and an accumulator for contact times

[pm-blanco]

Added:

• observables.PairwiseDistances: an observable that tracks pairwise distances between two sets of particles.
• accumulators.ContactTimes: an accumulator that tracks contact times within contact threshold for a set of input distances.
• samples\contact_time.py: a sample script showcasing how to use these new features.

The contact time is defined as:

$$\tau = t_f - t_0$$

where $t_f$ is the last measured time at which those particles were in contact and $t_0$ the first measured time at which those particles were in contact. For example, if two particles have been only in contact during two consecutive configuration then their contact time is $\tau = dt$, where $dt$ is the time step. If they are only in contact for a single configuration then their contact time is $\tau = 0$.

[RudolfWeeber]

Thank you Pablo. This is a very useful addition.
IMO, we should try to split this into

• an observable PairwiseDistance(pids1,pids2)
• and an accumulator ParticlecontactTimes(pairwise_distance_obs, contact_distance)

In my impression, this would allow us to leave the observabl genuinely stateless (and avoid cirumenting teh them being const). All the state would be in the accumulator.

Should we do a short video call, to discuss?

[pm-blanco]

Hi Rudolf! The solution you propose makes sense to me, the "hack" I had to do felt a bit against the current phylosophy of the observable/accumulator framework. I am simply not too familiar with the c++ core of the accumulators/observables so I did not want to tinker with it too much. I think I got myself somewhat familiar now with the observable framework, so I can try to refactor the feature to match what you propose.

Regarding the videocall, sounds good. I will send you an e-mail so we can schedule a time 😄

[pm-blanco]

@jngrad I have the PR mostly ready for review, but my test is skipped by the CI for some reason. Do you know why?

The following tests did not run:
	 30 - contact_time (Skipped)

[jngrad]

@jngrad I have the PR mostly ready for review, but my test is skipped by the CI for some reason. Do you know why?

The following tests did not run:
	 30 - contact_time (Skipped)

The class is missing a test_something(self) method to check the result of the simulation.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[jngrad]

@pm-blanco I think something went wrong during the rebase. Could you please try this?

git reset --soft upstream/python # rewind changes accidentally taken from the python branch
git commit -m 'Add contact time observable'
git show # here carefully inspect the diff to make sure it only contains your feature
git push -f

If this doesn't help, use git reflog to look through the history of orphaned commits and resurrect e.g. a38417ee4.