Network Graph, Chord & Arc Diagram Maker
An edge list becomes a force-directed graph, chord and arc diagrams and ranked metrics.
Try before you buy.
- Free preview: the graph with a watermark, the summary counts and the top rows of the metrics table by PageRank (up to 10).
- Locked until you unlock it: download.
- Unlock: Pro pass, ₹179 for 30 days or ₹79 for 1 week, a one-time payment that never renews. Unlock: Pro pass, $5.99 for 30 days or $2.99 for 1 week, a one-time payment that never renews.
Ways to unlock shows how to get the full result.
Printing this result is locked in the free preview.
One edge per line: source, target and optionally a weight and a type (commas, tabs or semicolons), or “A -> B 3”.
Node table (optional)
Id, then a label, a group and a size if you have them. Ids not in the edges become nodes on their own.
Look
Drawing
Selected node
Select a node in the drawing, find it by name, or choose one in the table.
Metrics
Locked in the free preview. Opens the ways to unlock this result.
Locked in the free preview. Batch runs unlock with a pass.
Locked in the free preview. Query results unlock with a pass.
About the Network Graph, Chord & Arc Diagram Maker
Paste an edge list — one connection per line, as “source, target, weight” or “A -> B” — and, if you have one, a node table with labels, groups and sizes, and see the network three ways: a force-directed graph that settles as you watch, a chord diagram of the flows between groups, and an arc diagram of the nodes in a row. Search for a node, highlight its neighbours, hide light edges with the weight filter, and colour or size the nodes by group, community or any metric.
For every node the page works out its degree and weighted degree (in and out for directed graphs), betweenness and PageRank, and finds communities with the Louvain method — all in your browser, in a background worker, for networks of up to 10,000 nodes. A ranked table lists them, ready for a spreadsheet or for Gephi.
How to use it
- Paste the edge list into Edges — source and target on each line, then a weight and a type if you have them — or open a CSV or Excel file, or pick an example. Tick Directed if each edge goes one way.
- Optionally paste a node table (id, label, group, size) to name the nodes and colour them by group.
- Choose the view — Network, Chord or Arc — and what colours and sizes the nodes; drag to pan, scroll or pinch to zoom, and select a node (or search for it) to see its neighbours and metrics.
- Download the drawing as PNG or SVG, the metrics as CSV or Excel, or the graph as GEXF or GraphML for Gephi — with a Pro pass, or after unlocking this graph; without one the page shows a free preview.
Examples
Source, Target, Weight, Type Asha, Ben, 5, project Ben, Chen, 4, project Chen, Dara, 2, review
Ben and Chen have degree 2; Ben’s weighted degree is 9; Ben and Chen lie on every path from Asha to Dara, so they have the highest betweenness.
34 club members, 78 friendships (an example on this page)
Member 1 has the highest betweenness (0.438) and member 34 the highest PageRank (0.101); Louvain finds about four communities (modularity above 0.4) that follow the club’s split.
The network from W. W. Zachary’s study of a karate club that split in two, the classic test of community detection.
app -> router app -> store router -> store store -> api
store has in-degree 2 and out-degree 1; with Directed ticked, the arc diagram draws edges going right above the line and edges going left below it.
Common uses
- Social network analysis: who connects groups (betweenness), who is most central (PageRank), which circles exist (communities).
- Flows between regions, departments or accounts as a chord diagram.
- Software dependencies, citation networks, co-authorship, supply chains and knowledge graphs.
- Preparing a network for Gephi, Cytoscape or NetworkX, with its metrics already worked out.
How the metrics are worked out
- Degree: the number of edges a node has (repeated edges between the same pair count once, their weights added). On a directed graph, in-degree and out-degree count incoming and outgoing edges.
- Weighted degree: the sum of the weights of a node’s edges.
- Betweenness: the share of shortest paths between other pairs of nodes that pass through a node, by Ulrik Brandes’ algorithm on unweighted paths, normalised by (n − 1)(n − 2) so it runs from 0 to 1 — the same values NetworkX gives. Above 3,000 nodes it is estimated from 1,000 sources spread over the graph, and the page says so.
- PageRank: the chance that a random walker — following edges in proportion to their weight, and jumping anywhere 15% of the time (damping 0.85) — is at a node. The values add up to 1.
- Communities: the Louvain method of Blondel, Guillaume, Lambiotte and Lefebvre moves nodes between communities while the modularity rises, then merges each community into one node and repeats. Directed edges count in both directions here. Community 1 is the largest.
The three views
The network is laid out by a force simulation in the manner of d3-force: edges pull their nodes together like springs, all nodes push each other apart (computed with a Barnes–Hut tree, so large graphs stay quick), and a weak pull keeps separate parts in view. The same graph always gets the same drawing. Node areas follow the metric you choose; edges are wider with more weight; labels show for the largest nodes that have room.
The chord diagram totals the edge weights between groups — the node table’s groups, or the communities — and draws each group as an arc as long as its total, joined to the others by ribbons whose ends show the flow each way. The arc diagram puts the nodes on a line, sorted by group and degree (or as you choose), and draws each edge as an arc; on a directed graph, edges going right sit above the line and edges going left below.
The files
- PNG and SVG of the view on screen (the SVG keeps every node, edge and label as a shape).
- CSV and Excel of the metrics table: rank, id, label, group, community, degrees, weighted degrees, betweenness and PageRank; the Excel workbook adds a sheet of the edges. In the CSV, cells that a spreadsheet would run as formulas start with an apostrophe; the workbook stores ids, labels and groups as text.
- GEXF 1.3 for Gephi, with every metric as a node attribute, the edge weights and types, and the drawing’s positions, colours and sizes, so Gephi opens the graph as you see it here.
- GraphML with the same attributes as typed keys, for Cytoscape, yEd, NetworkX and igraph.
Limitations
- Up to 10,000 nodes and 100,000 edges. The network is drawn on a 2D canvas, so very large graphs pan and zoom slowly on phones.
- Betweenness uses unweighted shortest paths (the number of steps), and is estimated above 3,000 nodes.
- Louvain communities depend a little on the order of the nodes; with modularity this close between different splits, a reordered list can give a slightly different but equally good answer.
- Nodes cannot be dragged into place; the layout is the simulation’s own.
Privacy
On this device: nothing is sent
Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot.
Frequently asked questions
What do I get without a pass?
Without a pass, Network Graph, Chord & Arc Diagram Maker shows the graph with a watermark, the summary counts and the top rows of the metrics table by PageRank (up to 10). Until you unlock it, the result can’t be downloaded. A Pro, Premium or Ultimate pass, a one-time payment that never renews, unlocks the full result. The pricing page lists the passes and their prices.
What format should my edge list be in?
One edge per line: the source, the target and optionally a weight and a type, separated by commas, tabs or semicolons — copied from a spreadsheet works. A first line such as “Source, Target, Weight” is recognised. Lines like “A -> B” or “A -> B 3” work too.
What do betweenness and PageRank tell me?
Betweenness is high for nodes that sit on the shortest paths between others — brokers and bridges between groups. PageRank is high for nodes that many well-connected nodes point to (or are linked with) — the most central or influential ones.
Can I open the result in Gephi?
Yes: the GEXF file carries the metrics, the communities, the positions, the colours and the sizes, so Gephi shows the same picture and can run its own statistics on top. GraphML works in Gephi too.
Why are my edges merged?
Two lines with the same two nodes describe one connection, so they become one edge with the weights added (on a directed graph, only lines in the same direction merge). Lines from a node to itself are left out. The page says how many.
Is my network uploaded?
No. Reading, metrics, communities and the layout all run in your browser (in a background worker), and the files are made on your device.