Improve Logging System & Enhance FIRE Algorithm for Compound Molecules

[LMSM3]

Overview

The project already includes a light ML automation pipeline. This issue will focus on two improvements:

1. Better Logging System

   • Design and implement a more robust logging system for the pipeline and simulation steps.
   • Aim: Improved traceability, easier debugging, and actionable logs for automation steps.
   • Suggested: Consider integrating structured logging (e. g., JSON logs, timestamps, levels) and optional file-based/system log output.

2. Improving FIRE for Compound Molecules

   • Strengthen the Fast Inertial Relaxation Engine (FIRE) algorithm particularly for handling compound (multi-component) molecules.
   • Evaluate current limitations or edge cases where compound geometries and optimizations fail or are suboptimal.
   • Possible approaches: enhanced convergence criteria, special-case handling for multi-center atoms, user-configurable parameters for compound types.

Acceptance Criteria

• Robust logging established for all automation steps and ML submodules.
• FIRE improvements validated with compound/multi-component molecular input, with before/after benchmarks or examples if possible.
• Documentation updated for new logging options and usage notes for enhanced FIRE support.

---

Further suggestions, benchmarks, or implementation notes are welcome in comments.