Coherence Law: Minimal Dataset and Code for Testing Δ–log E Relation

This dataset contains a minimal reproducible test kit to evaluate the proposed linear relation:  \log E \approx k\cdot\Delta + b  where: is a local path-information gap (e.g., mutual information difference under surrogate disruption), is an endurance horizon (e.g., time-to-threshold under noise, or inverse Lyapunov rate). Included: README.md: Explanation of the relation, variable definitions, falsifiability checklist. data.csv: Toy dataset showing values of , , and domain grouping. coherence_test.py: Simple Python script that runs linear regression, plots the result, and outputs fit metrics. This archive is intended to serve as a clear and falsifiable entry point for independent testing. It requires only common Python packages (pandas, numpy, sklearn, matplotlib) and runs in seconds. For further theoretical and applied context, see companion reports: Zenodo:17145179