Dataset1 / Dataset2 source trace
Original DA S1/S2 code, split by source family.
This atlas separates the extra course Python files from Sector4. Dataset1 covers CPV/ATR demos and local repair scripts. Dataset2 covers index strategy demos, feature generation, XGBoost, symbolic regression and the larger gplearn_numba engine.
dataset1 / dataset2 / gplearn
module trace -> public summary
01
Source Families
02
Python Files
Claim Boundary
Extra course source, not the cleaned Future OS.
Original DA S12 is a source-reading atlas. It keeps Dataset1 and Dataset2 code separate from Original Sector4 so the refactor story stays clean: raw course code with hard-coded local paths on one side, engineered OS modules with config and loaders on the other side.