Add ASV benchmarks for MDAnalysis.lib core kernels

[Amarendra22]

Summary

This PR adds ASV benchmarks for several core computational kernels in MDAnalysis.lib.

The following functions are benchmarked:

• distance_array
• self_distance_array
• calc_dihedrals
• FastNS neighbor search
• AtomNeighborSearch

These functions are widely used across MDAnalysis analysis modules and are performance-critical components. Adding benchmarks helps track performance regressions and evaluate future optimizations.

Benchmarks added

• lib.distances.DistanceArrayBench.time_distance_array
• lib.distances.SelfDistanceArrayBench.time_self_distance_array
• lib.dihedrals.DihedralBench.time_calc_dihedrals
• lib.fastns.FastNSBench.time_fastns
• lib.neighborsearch.NeighborSearchBench.time_neighbor_search

Motivation

The MDAnalysis.lib module contains low-level computational kernels that are called repeatedly by higher-level analysis tools. Benchmarking these functions improves performance monitoring and supports future optimization work.

LLM / AI generated code disclosure

LLMs or other AI-powered tools were used in this contribution: No (used for research purpose only)

)

Developers Certificate of Origin

I certify that I can submit this code contribution as described in the Developer Certificate of Origin, under the MDAnalysis LICENSE.

---

📚 Documentation preview 📚: https://mdanalysis--5292.org.readthedocs.build/en/5292/

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.83%. Comparing base (6e837a4) to head (20607e7).

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #5292      +/-   ##
===========================================
- Coverage    93.83%   93.83%   -0.01%     
===========================================
  Files          182      182              
  Lines        22501    22501              
  Branches      3195     3195              
===========================================
- Hits         21114    21113       -1     
- Misses         924      925       +1     
  Partials       463      463              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[orbeckst]

There may be some overlap with potential GSOC projects so I am going to block this PR for the time being.

[orbeckst]

For all the distance calculation, note that these functions can also work with periodic boundary conditions (PBC, box=... kwarg) and for these boxes it matters if they are orthorhombic or triclinic. The distances should be commensurable with the box sizes.

For the dihedrals it would be nice to not use totally random numbers but generate 4-tuples of atoms for random dihedral angles first.

We also have different backends (see docs).

We would like to parametrize over

• box = None, orthorhombic shape, triclinic shape
• backend = serial, OpenMP, distopia

[orbeckst]

There's already a contacts benchmark in develop. Remove this one here.

[orbeckst]

This should also automatically resolve the merge conflict that is indicated in the GitHub status checks box.

[orbeckst]

This file may not be needed. Remove it and see it it still works.