Indexing the contributor, not the byline: a contribution-weighted logarithmic index and its empirical comparison with the h-index

Standard citation metrics assign each paper's full citation count to every co-author. This project implements and evaluates an alternative: an index that partitions each publication's log-citation credit among its authors according to contribution shares, P(a) = Σ s·ln(1+c), alongside its linear companion C(a) = Σ s·c. Because declared contribution shares do not exist retrospectively, the pipelines compute both indices under three proxy weighting schemes (uniform fractional, harmonic, first–last) and compare the resulting rankings with the h-index. The repository provides: (1) reproducible Python pipelines against the open OpenAlex database (CC0) and, for cross-database replication, the Scopus API; (2) an identity-verification workflow with automatic flags for profile fragmentation, namesake merging, and low-confidence name matches; (3) summary data for three cohorts 15 researchers in plant science, 13 researchers across ten disciplines, and the resolvable segment (n = 213) of a national top-500 author list; and (4) the citation inflation factor, a scheme-free diagnostic of consortium-dominated publication portfolios. The indices describe the structure of recorded contribution; they measure neither the quality nor the integrity of any researcher's work. See README interpretation notes before reusing the data.