- Updated script to generate SHAP beeswarm plots for both the RBF objective and target surrogates.
- Refactored model loading functions for clarity and efficiency.
- Added functionality to skip objective plots via command line argument.
- Improved error handling for missing model files and data.
- Streamlined SHAP value computation and output saving processes.

Refine cross-validation strategy descriptions and improve manuscript clarity in the methodology section

Improve manuscript clarity and formatting in the section on buckling-delayed shear-link dampers. Adjust phrasing for consistency and enhance readability in various sections.

Remove the manuscript PDF for the adaptive FEM-validated surrogate optimization of buckling-delayed shear-link dampers.

Fix typos and improve clarity in the manuscript on surrogate optimization for optimal design layout. Update variable names for consistency and correct formatting in equations. Update binary file for synchronization.

- Revised the abstract and introduction sections of the manuscript to clarify the optimization framework for buckling-delayed shear-link dampers.
- Updated design variables and parameters in the manuscript to improve clarity and accuracy.
- Enhanced the validation section with additional details on the FEM model and experimental setup.
- Improved the damage-aware optimization methodology, emphasizing the balance between damage control and energy dissipation.
- Added a new reference for a conference paper on nonlinear dynamic analysis of structures equipped with shear links.
- Introduced a new PDF document related to the optimization layout.

This commit introduces a new script, `rbf_objective_shap_beeswarm.py`, which generates a SHAP beeswarm plot for the RBF optimization objective function. The script includes functionality to load RBF models, prepare background data, compute SHAP values, and save the results in various formats. It also implements a uniform design for sampling and applies a specific styling for the plots.

- Replaced specific geometry family names (H30_B29, H30_B34, etc.) with generic identifiers (F_1, F_2, etc.) throughout the manuscript for confidentiality.
- Adjusted descriptions and captions in figures and tables to reflect the new identifiers.
- Updated design variable definitions and thickness ranges in the manuscript to align with the new naming convention.
- Enhanced clarity in the discussion of optimization results and surrogate model comparisons.
- Ensured consistency in the presentation of data across all sections of the manuscript.

- Updated the RESILINK surrogate optimization manuscript PDF to reflect recent changes.
- Added a new synctex.gz file for improved LaTeX editing and referencing.
- Revised the manuscript text to include conclusions in the workflow summary.
- Introduced a new section on dataset generation and surrogate modeling, detailing the use of FEM models and damage indicators.
- Removed redundant section header for clarity in the design of experiments description.

- Created new manuscript file for RESILINK surrogate optimization.
- Added PDF version of the manuscript.
- Updated LaTeX file to reference the correct journal name and include necessary figures.
- Adjusted figure paths to ensure correct referencing in the document.
- Modified USG.cls to correctly handle page numbering for the manuscript.
- Updated bibliography path in the manuscript for proper citation.

- Clarified the description of FEM simulations and loading patterns based on device height.
- Improved the explanation of damage indicators and their computation for optimization.
- Updated the definitions of local shear distortion and its role in energy dissipation.
- Enhanced the description of supervised ML surrogate models and their training process.
- Added detailed hyperparameter search spaces for various models used in Bayesian optimization.
- Revised the objective function formulation to emphasize damage control and energy dissipation balance.
- Improved clarity in the penalty definitions for window and frame damage.
- Summarized the validation process for optimized geometries and the iterative nature of the optimization loop.

Refine manuscript text for clarity and consistency in the description of design variables, optimization cycles, and surrogate model performance. Adjust punctuation and phrasing for improved readability throughout the document.

- Added new figures: TempBayesian.png and TempBayesian.svg to the Figures directory.
- Updated references in the manuscript to use PDF versions of figures for better quality:
  - MethodologyFlowChart.png to MethodologyFlowChart.pdf
  - BayesianSearchCV.png to BayesianSearchCV.pdf

- Updated manuscript title to reflect a focus on comparison of surrogate strategies for damage-aware optimization.
- Improved clarity in the results section by refining descriptions of surrogate model performance and validation metrics.
- Enhanced the discussion on the adaptive validation loop and its impact on optimization outcomes.
- Clarified the interpretation of objective-function values and the implications of damage control in the optimization framework.
- Added figures to illustrate the evolution of optimized window thicknesses and RBF objective surfaces during the optimization process.

- Created `surrogate_selection_summary_by_variable_type.csv` and `surrogate_selection_summary_by_variable_type_wide.csv` to summarize model performance across different variable types and window groups.
- Updated the manuscript to include new figures and tables reflecting the results of the surrogate model comparisons, including predictive performance and optimization results.
- Adjusted figure placements to improve layout and readability.
- Enhanced discussions on the effectiveness of supervised ML models versus RBF interpolation in the context of surrogate-assisted optimization.

Update manuscript to enhance clarity in RBF surrogate validation and streamline adaptive FEM validation section

- Updated the objective function to use area factors instead of volume factors for window contributions.
- Clarified the formulation of damage penalties, emphasizing the severity of frame damage.
- Adjusted terminology in the manuscript to reflect changes in the optimization approach.
- Revised table captions and contents to align with the new area factor usage.

- Introduced a new Python script `plot_load_patterns.py` to visualize loading pattern displacement histories from `LoadPatterns.csv`, generating figures in PNG, SVG, and PDF formats.
- Removed the outdated `PatronesCarga.csv` file as it is no longer needed.
- Revised the abstract and various sections of the manuscript to enhance clarity and detail, including the introduction of radial basis function surrogates and adjustments to the optimization methodology.
- Updated references in the bibliography to include new citations for SciPy and Differential Evolution.