What problem does it solve? Fitting 3D curves (such as log-spirals) to anatomical point clouds and rebuilding them into figures is error-prone: missing center offsets anchor curves to the wrong origin, argsort-based ordering scrambles spiral paths, and inconsistent scripts produce conflicting RMSE values. This Skill codifies the correct fitting pipeline and its known pitfalls so figures are accurate and publication-ready. ## Core Features & Use Cases - Correct 3D Reconstruction: Enforces rebuild formulas that explicitly include the spiral center offset (cx, cy) instead of anchoring curves to the data centroid. - Path-Based Fitting: Requires nearest-neighbor path ordering with np.unwrap for angle expansion, forbidding np.argsort(theta) in final parameter fitting. - Cross-Validation & Consistency: Verifies RMSE, spiral rate b, and visual quality together, and checks that multiple figure scripts output identical parameters. - Use Case: When fitting a log-spiral to cochlear canal centerlines from micro-CT or MRI segmentations, use this Skill to merge two-ended segment annotations, fit the 8-parameter model, and generate multi-specimen composite figures with verified RMSE labels. ## Quick Start Fit a 3D log-spiral to my AC bony point cloud from the micro-CT specimen and generate a publication-grade figure with RMSE and parameter annotations.
