Proximity analysis
Compare every selected manuscript with every other manuscript. Exclusive/common agreement controls are deliberately absent here: the output is pairwise proximity across the whole selected group.
Loading the apparatus and Vīdēvdād datasets…
Run the analysis to calculate the matrix.
Run the analysis before displaying the network.
Select a matrix cell or network edge to inspect a manuscript pair.
Bayesian affinity ρ. This is a provisional stemmatic affinity measure, not a raw identity percentage. Shared rare readings contribute more than readings that are widespread in the comparison population. The posterior uses a Beta(1,9) regularizing prior and reports a 95% credible interval in pair details. In apparatus mode, explicit var. has full weight; included reading groups have lower weight. In transliteration mode, exact aligned agreements are discounted by their frequency across the selected Vīdēvdād witnesses.