TransportNetworkWelfare.jl

This package evaluates local transport improvements in spatial equilibrium. A project consists of two CSV files and one TOML file. The computation returns directed-arc effects, physical-link effects, closure decompositions, parameter sensitivity paths, and diagnostics.

For a paper-centered treatment of the theory and a complete computational walkthrough, build the practitioner guide with make practitioner-guide from the repository root. Use make practitioner-guide-check to verify its numerical assets, citations, references, and PDF.

Install and run the toy project

julia --project=. -e 'using Pkg; Pkg.instantiate()'
julia --project=. bin/tnw.jl validate examples/toy/config.toml
julia --project=. bin/tnw.jl decompose examples/toy/config.toml

The first command installs Julia dependencies. The next two commands need no external data. To adapt the example, change only nodes.csv, edge_modes.csv, and config.toml; source edits are not part of the workflow.

For a separate project, initialize a portable template and point the package CLI to its configuration:

julia --project=. bin/tnw.jl init /path/to/my-network
julia --project=. bin/tnw.jl validate /path/to/my-network/config.toml
julia --project=. bin/tnw.jl decompose /path/to/my-network/config.toml

The own-data guide states the required accounting identities and explains which raw transport measures need application-specific preprocessing.

Output contract

Each run writes CSV results, diagnostics, and run_manifest.json. The manifest records the package version and commit, Julia version, command, configuration hash, input hashes, transformations, model variant, parameters, condition numbers, verification status, and output hashes.

The code repository owns source, tests, configurations, and small synthetic fixtures. The journal manuscript remains separate. Restricted data remain outside Git and are resolved through environment variables. External data and derived assets retain their source terms, as described in THIRD_PARTY_NOTICES.md.