Subject
2 entries
Causality
Bookmarks
SURD: Synergistic-Unique-Redundant Decomposition of Causality
SURD is a Python library for decomposing causality into synergistic, unique, and redundant components in complex systems — quantifying how different variables jointly contribute to predicting future states. Developed for turbulence analysis but applicable to any causal inference problem.
Simpson's Paradox: Interactive Visualization
VUDLab's interactive visualization of Simpson's Paradox — showing how a trend that appears in every subgroup can reverse when the groups are combined. A crucial statistical phenomenon for anyone working with observational data.