DeLPHI v1.0.0: trained networks, simulated data, predictions, results and inversion fits for asteroid pole-axis candidates
收藏资源简介:
Data for the revised DeLPHI Paper I, used with software release v1.0.0. This record brings together the trained networks, synthetic data and original evidence with the revision results: estimated periods, predictions, score maps, controls, reliability and external-photometry tests, spacecraft comparisons, and inversion summaries. Three new archives provide the own-period results, saved per-start convexinv benchmark fits at estimated and DAMIT periods, and additional cited supporting files. The README maps the manuscript's sections and Appendices A–E to the files, explains extraction and DELPHI_DATA, and gives the acquisition code's addresses for third-party inputs. The networks were trained on DAMIT-derived data. The new archives exclude raw third-party downloads and prepared photometry; the unchanged historical evidence archive retains its small DAMIT census export. Third-party material keeps its original terms. Our results are licensed CC BY 4.0; code retains its software license. Software: DeLPHI v1.0.0. Data DOI: 10.5281/zenodo.23757942.



