How sjkoelle/isometry-pursuit works
Key Components to Include in the System Design:
Core Algorithm Components (Central Module)
• algorithm.py (Main algorithm implementation)
• atomic_geometry.py (Geometric computations)
• loss.py (Loss function calculations)
• manifold.py (Manifold operations)
• transformation.py (Transformation utilities)
Supporting Components (Secondary Layer)
• experiment.py (Experiment framework)
• simulation.py (Simulation utilities)
• plot.py (Visualization tools)
• vendor.py (Third-party integrations)
Experimental Framework (Outer Layer)
• Experiments/ directory components:
• Jupyter notebooks (.ipynb files)
• Results storage (CSV files)
• Standardized datasets (iris and wine data)
Documentation Layer
• Workshoppaper/ component showing the academic documentation
Design Guidelines:
Component Organization
• Use a hierarchical layout with core components at the center
• Rectangles: Core processing modules
• Rounded rectangles: Supporting utilities
• Documents: Documentation components
Supporting Layer
• Experiment.py, simulation.py, plot.py, and vendor.py in an outer circle
Experimental Framework
• Jupyter notebooks together
• Data flow between experiments and results
• To core components through experiment.py
Documentation Layer
• Workshoppaper/ components on the side
Level of Detail:
• Keep implementation details within modules hidden
• Component interactions and data flow
• Algorithmic steps without diving into mathematical details
• Clear separation between experimental validation and core algorithm
Additional Considerations:
2. Version control (.git) and project configuration (setup.py) as administrative components