EgoLines

Instructions for file uploading

EgoLines accepts a simple JSON file or two CSV files (a node list and an edge list) for storing a dynamic ego-network.

1) JSON format

Here is an example JSON file.

2) CSV format

The node list contains at least two columns (in such order): time index, and node id. Additional columns can be appended for node attributes.

The 1st row of the CSV file must be the header.
The 2nd row must be the meta-data, representing: directed network or not ("T" or "F", the 1st column), node id of the ego (the 2nd column), and types of node attributes (if any, "categorical" or "numerical", starting from the 3rd column).
The real data begins from the 3rd row, and the data rows must be sorted by their time indices.

The edge list contains four columns (in such order): time index, source node id, target node id, and weight.

The 1st row of the CSV file must be the header.
The real data begins at the 2nd row, sorted by time indices.

Here are example node CSV file and edge CSV file.

Further notes

There are a few things to note about:

Node attributes are optional, and if none is provided, node degrees will be computed and added as an attribute.
If the input is a directed network, the order of the source and target nodes of an edge does not matter, and thus there is no need to include node1->node2 and node2->node1 at the same time (which are the same edge).
If the input is an unweighted graph, set the edge weights all to 1.
Currently EgoLines best supports ego-networks up to 2 levels, i.e., networks including the ego, its immediate neighbors, and its neighbors' neighbors. However, EgoLines is capable of displaying ego-networks with more than 2 levels or general networks, but in this case the "zone" tool is limited to just highlight immediate neighbors of a chosen (ego) node.

Also note that we never store your file on any server. All the file processing happens on the client side and will be deleted when you clear your browser cache.