External Corpora Preparation
Generated by scripts/prepare_external_corpora.py.
The prepared matrices reserve numeric state 0 for missing readings. Positive integers are per-feature categorical state codes; state-code mappings are under outputs/tables/external_corpora/.
| Dataset | Rows | Raw features | One-hot indicators | Missing fraction | Target |
|---|---|---|---|---|---|
catena_ephesians |
267 | 1206 | 3001 | 0.571 | full_matrix_average_linkage |
catena_galatians |
393 | 1314 | 3399 | 0.545 | full_matrix_average_linkage |
catena_romans |
340 | 3589 | 8707 | 0.647 | full_matrix_average_linkage |
comp_tc_mark_stemma_backbone |
184 | 148 | 790 | 0.379 | non_red_dot_forest_backbone |
Target notes
catena_ephesians
Machine-readable fallback target from average-linkage simple-matching distance over the full NEXUS matrix; the published MrBayes tree is preserved as data/raw/external/catena/ephesians/Ephesians-phylogenetic-tree_10May24.pdf but not machine-parsed.
- Numeric feature file:
data/processed/catena_ephesians__features_numeric.csv - Source: https://edata.bham.ac.uk/1110/
catena_galatians
Machine-readable fallback target from average-linkage simple-matching distance over the full NEXUS matrix; the published MrBayes tree is preserved as data/raw/external/catena/galatians/Galatians-phylogenetic-tree_10May24.pdf but not machine-parsed.
- Numeric feature file:
data/processed/catena_galatians__features_numeric.csv - Source: https://edata.bham.ac.uk/1108/
catena_romans
Machine-readable fallback target from average-linkage simple-matching distance over the full NEXUS matrix; the published MrBayes tree is preserved as data/raw/external/catena/romans/Romans-phylogenetic-tree_10May24.pdf but not machine-parsed.
- Numeric feature file:
data/processed/catena_romans__features_numeric.csv - Source: https://edata.bham.ac.uk/1112/
comp_tc_mark_stemma_backbone
Graphviz Stemmaweb DOT projected by dropping red reticulate edges and attaching the resulting forest under a synthetic root.
- Numeric feature file:
data/processed/comp_tc_mark_stemma_backbone__features_numeric.csv - Source: https://github.com/jjmccollum/comp-tc-mark