menobis filter
TL;DR
Use menobis filter to classify observed edges as upper-significant,
lower-significant, or compatible with an independent MENoBiS null model.
Commands
| Command | Null model | Input |
|---|---|---|
strength-poisson |
fit strength Poisson | edge table |
strength-cost-poisson |
fit strength-cost Poisson | edge table + coordinates CSV |
strength-edges-poisson |
fit strength-edges Poisson (zero-inflated) | edge table |
strength-degree-poisson |
fit strength-degree Poisson (zero-inflated) | edge table |
degree-events-poisson |
fit degree-events Poisson (zero-inflated) | edge table |
custom-poisson |
user-supplied Poisson rates | edge table + rates CSV |
Examples
menobis filter strength-poisson edges.csv --output-prefix filtered/
menobis filter strength-edges-poisson edges.csv --target-edges 500 --output-prefix filtered/
menobis filter strength-cost-poisson edges.csv --coordinates xy.csv --target-cost 1.5 --output-prefix filtered/
menobis filter strength-degree-poisson edges.csv --output-prefix filtered/
menobis filter degree-events-poisson edges.csv --output-prefix filtered/
menobis filter custom-poisson edges.csv --rates rates.csv --output-prefix filtered/
Outputs
--output-prefix DIR writes:
| File | Meaning |
|---|---|
upper.csv |
observed edges heavier than expected |
lower.csv |
observed positive edges lighter than expected |
compatible.csv |
observed edges compatible with the null |
absent_lower.csv |
absent pairs that should exist under the null |
Each file contains p-values, expected occupation, and occupation probability.
Options
| Option | Meaning |
|---|---|
--alpha |
significance level |
--tail |
upper, lower, or two-sided |
--correction |
none, bonferroni, or fdr |
--detect-absent |
stream absent-pair candidates |
--min-occupation |
absent-edge occupation threshold |
--min-expected |
absent expected-weight threshold |
--max-absent |
cap absent output size |
--self-loops/--no-self-loops |
diagonal handling |
--target-edges |
target edge count (strength-edges-poisson) |
--target-cost |
target total cost \(\sum t_{ij}d_{ij}\) (strength-cost-poisson) |
--coordinates |
Projected XY coordinate CSV path (strength-cost-poisson) |
--rates |
rates CSV path (custom-poisson) |