CLI Reference

ForceAtlas2 provides a command-line interface via python -m fa2.

Commands

layout — Compute positions

python -m fa2 layout edges.json -o layout.json
python -m fa2 layout edges.csv --mode community --iterations 200 --seed 42
cat edges.json | python -m fa2 layout - > positions.json

render — Layout and render to image

python -m fa2 render edges.json -o graph.png
python -m fa2 render edges.csv -o graph.svg --title "My Network" --mode community

metrics — Compute layout quality

python -m fa2 metrics edges.json
python -m fa2 metrics edges.json --positions layout.json

Options

Flag	Short	Default	Description
--iterations	-i	100	Number of layout iterations
--dim	-d	2	Layout dimensions
--mode	-m	default	Layout preset: default, community, hub-dissuade, compact
--seed	-s	None	Random seed
--output	-o	stdout	Output file path
--title	-t	None	Chart title (render only)
--positions	-p	None	Pre-computed positions file (metrics only)

Input Formats

JSON

[["A", "B"], ["B", "C", 5.0]]

Or with explicit structure:

{"edges": [["A", "B"], ["B", "C"]], "nodes": ["A", "B", "C"]}

Or adjacency dict:

{"A": ["B", "C"], "B": ["A"]}

CSV

source,target,weight
A,B,1.0
B,C,2.0

Piping

# Stdin → layout
echo '[["A","B"],["B","C"],["A","C"]]' | python -m fa2 layout - > layout.json

# Stdin → render
echo '[["A","B"],["B","C"],["A","C"]]' | python -m fa2 render - -o graph.png

# Stdin → metrics (computes layout internally)
echo '[["A","B"],["B","C"],["A","C"]]' | python -m fa2 metrics -