Networked resilience at a distance: replication package (data and code)

Data and code accompanying the article "Networked resilience at a distance: how the scientific diaspora sustains Ukrainian higher education in wartime — and why military education remains disconnected" (submitted to Humanities and Social Sciences Communications). The package contains a pseudonymised extract of the public scholars' registry of the Ukrainian Science Diaspora initiative (n = 92), the 74-node term co-occurrence network derived from a 12,101-record Scopus corpus (2012–2026), the VOSviewer thesaurus, percolation curves, intermediate computation values, and an end-to-end replication script (compute_all.py; Python 3.12, networkx 3.6, fixed seeds) reproducing every reported network metric. See README.md for file descriptions.