Supplementary material for "Lot Planning as an Input Transformation for Scheduling Methods": instances, validation results and oracle code

Supplementary material for the article "Lot Planning as an Input Transformation for Scheduling Methods". The archive contains the test instances (S1), the validation results reported in Section 4 of the article (S2), and the independent oracle and verification code (S3). The verification program S3_Code/verify_supplementary.py re-checks the published results without any part of the planning system: it recomputes the lot-sizing optimum of the worked example, the optimal sublot sizes, the batching rules, the rule-gap table and the lot-streaming table directly from the raw data, using only the independent oracle in oracle.py. It requires only pandas and openpyxl, and prints ALL VERIFIED when every check passes. S3_Code/run_validation.py is the full validation suite of 59 checks reported in the article. It exercises the planning system itself against the oracle and therefore has to be run from inside the system's source tree; it is included so that the protocol is documented in full.