ClusteringMethods don't need to return details

[lilyminium]

Expected behavior

Everything returned by ClusteringMethod.__call__ contains information that is used.

Actual behavior

ClusteringMethod.__call__ returns clusters, details. details is always an empty dictionary. It's not used by cluster and is not even in the documentation:

mdanalysis/package/MDAnalysis/analysis/encore/clustering/ClusteringMethod.py

Lines 77 to 91 in 9bcf6f4

	def __call__(self, x):
	"""
	Parameters
	----------
	x
	either trajectory coordinate data (np.array) or an
	encore.utils.TriangularMatrix, encoding the conformational
	distance matrix
	Returns
	-------
	numpy.array
	list of cluster indices
	"""

The documentation is also inaccurate: clusters is an array of the centroid frame indices of length #elements, as returned by encode_centroid_info:

mdanalysis/package/MDAnalysis/analysis/encore/clustering/ClusteringMethod.py

Lines 57 to 66 in 9bcf6f4

	def encode_centroid_info(clusters, cluster_centers_indices):
	"""
	Adjust cluster indices to include centroid information
	as described in documentation for ClusterCollection
	"""
	values, indices = np.unique(clusters, return_inverse=True)
	for c_center in cluster_centers_indices:
	if clusters[c_center] != c_center:
	values[indices[c_center]] = c_center
	return values[indices]

Code to reproduce the behavior

As shown below, cluster does not use the empty details dictionary.

mdanalysis/package/MDAnalysis/analysis/encore/clustering/cluster.py

Lines 234 to 235 in 9bcf6f4

	ccs = [ClusterCollection(clusters[1][0],
	metadata=metadata) for clusters in results]

Currently version of MDAnalysis

• Which version are you using? (run python -c "import MDAnalysis as mda; print(mda.__version__)") 0.20.2-dev
• Which version of Python (python -V)?
• Which operating system?

Suggested fix
Remove details from all the ClusteringMethods and update the documentation. Run the tests to make sure it doesn't affect anything else.

[shfrz]

Hi @lilyminium I would want to take this one!

[lilyminium]

@mtiberti are you happy with @shfrz fixing this issue?

[orbeckst]

@shfrz please feel free to work on this issue. @mtiberti can always review.

[shfrz]

@orbeckst sure, thank you!

[shfrz]

@MDAnalysis/gsoc-mentors could you please review my PR? Thanks!

[shfrz]

Hi, @lilyminium I implemented the suggested fix but 8 out of 12 tests are failing!

[kpiyush04]

Hi! @lilyminium @orbeckst
I want to work on this issue

[IAlibay]

@kpiyush04, there is already a PR by @shfrz aiming to deal with this issue (#2585). I would suggest possibly going for a separate GSOC starter issue as #2585 will be prioritised. If the PR stalls then it will be closed and we will look for alternatives.

[kpiyush04]

ok thanks a lot