CoNexDat / CoNexDat/LaNet-vi

Not ported: k-connectivity, Gomory-Hu edge connectivity, SVG/PDF output, -window cropping, weighted visuals

Open
#25 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
8
Forks
0
Avg merge
6h 10m
Merged PRs (30d)
16

Description

Found by the September 2026 comparison against the C++ sources (see docs/migration-from-cpp.md on branch docs/legacy-heritage (PR #17)). Tracking issue for whole features of LaNet-vi 3.0.1 with no Python equivalent; split into separate issues when scheduled.

- **k-connectivity** (`-kconn`, `-kconntype strict|wide`; `graph_kcores_components.cpp` computeKConnectivityWide/Strict): walks clusters from the top shell down, seeds with a cluster of contracted diameter ≤ 2 (wide also accepts min-cut ≥ shell), attaches clusters by frontier/φ conditions; non-k-connected nodes are recoloured (and drawn as cubes in bw) and `log/kconn.log` is written. This was a headline feature of the NJP 2008 paper and of "Understanding edge connectivity in the Internet through core decomposition" (Internet Mathematics 2011).
- **Gomory-Hu edge connectivity** (`-connectivity`, `-innerConnectivity`; `connectivity.cpp`, `ff_routine.cpp`): cut tree with unit-capacity Ford-Fulkerson, all-pairs connectivity aggregated by (min shell, flow), per-layer variant. NetworkX has `gomory_hu_tree`, so this is mostly glue plus the aggregation logs.
- **SVG output and PDF via rsvg** (`svg.cpp`, `svg_renderer.cpp`), **POV-Ray** (`povray.cpp`): matplotlib can save SVG/PDF, but the C++ scene semantics (fixed viewBox from u·2·gamma·R with 1.6:1.2 aspect, pale component fills, stroke = 2×diameter) are not reproduced.
- **`-window hstart hend vstart vend`** viewport cropping and exact `-W`/`-H` framing.
- **Weighted-graph visuals**: strength-based node radius, weight-based edge width, strength degree legend (`graphics_kcores.cpp`).
- **`-fromlayer`**: induced subgraph of index ≥ k, cores recomputed with the parent p-function, then the normal pipeline (tracked under the CLI/config issue as an inert option).
- Deliberately not ported (fine): `-java` viewer, `-onlygraphic`/`-nographic`, `-net` (never implemented in C++ either), `-coresfile` (dead in C++; the real artefact was `log/cores.log` in ` ` order).

Contributor guide

Open the contributing guide

Research direction

Start with docs/migration-from-cpp.md on branch docs/legacy-heritage and compare one listed feature with its named C++ sources, such as graph_kcores_components.cpp, connectivity.cpp, or svg_renderer.cpp. This tracking issue should first be narrowed to a single feature; done means its Python behavior, output, logs, and CLI options match the stated requirements, with focused validation added.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp, matplotlib, python
Domain
cli, data-visualization
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.