JSON API

Read-only access to the catalog for programs. All endpoints use GET, return JSON (except file downloads and drawings), and are relative to https://networks.skewed.de.

Parameters are given in the query string (e.g. ?tags=Social&full=true). Boolean parameters are false if they are left out, or given as 0, false, no or off (in any case), and true otherwise, also without a value (e.g. ?full).

List entries

GET /api/nets

Returns the names of all entries in the catalog, or their full records if full is given.

Parameter	Value	Description	Example
`full`	boolean	Return a dictionary mapping each name to its record (as in entry details) instead of a list of names.	`full=true`
`fields`	field names, separated by commas	With `full`, only these fields of the records (see entry details), which makes the response smaller and faster. Unknown fields respond with `400`.	`fields=nets,analyses`
`tags`	tag names, separated by commas	Only entries with at least one of the tags are returned. Tags are matched exactly (including their case); see the statistics for the tags in use.	`tags=Social,Offline`
`exclude`	entry names, separated by commas	Entries to leave out.	`exclude=advogato,chess`

Example: /api/nets?tags=Social,Offline

["7th_graders", "add_health", "ambassador", "baseball", ...]

Entry details

GET /api/net/<name>

Returns the record of a single entry. Responds with 404 if the entry does not exist.

Field	Description
`title`	Human-readable title.
`description`	Description, in Markdown.
`tags`	List of tags.
`date_added`	Date when the entry was added to the repository (`YYYY-MM-DD`).
`nets`	Names of the networks in the entry. For a single network, this is the entry name itself.
`analyses`	Summary statistics (`num_vertices`, `num_edges`, `average_degree`, …). For entries with several networks, a dictionary keyed by network name.
`citation`	List of `[text, url]` pairs for the upstream sources.
`bibtex`	BibTeX records for the citations.
`url`	Upstream URL.
`restricted`	Whether the files require authorization.

Example: /api/net/7th_graders

{
  "title": "Vickers 7th Graders (1981)",
  "tags": ["Social", "Offline", "Multilayer", "Unweighted", "Metadata"],
  "nets": ["7th_graders"],
  "analyses": {"num_vertices": 29, "num_edges": 740, "is_directed": true, ...},
  "restricted": false,
  ...
}

Download files

GET /net/<name>/files/<net>.<format>

Downloads a network file. <net> is one of the names listed in nets; for an entry with a single network it equals <name>. Files of restricted entries respond with 401 unless an access token, obtained on request, is given in the WWW-Authenticate header of the request, as Bearer <token> (this is what graph-tool does with the token it is given).

Format	Description
`gt.zst`	graph-tool binary format, zstd-compressed
`xml.zst`	GraphML, zstd-compressed
`gml.zst`	GML, zstd-compressed
`csv.zip`	CSV edge and node lists, zip archive

The largest networks (billions of edges) are only available in the gt.zst format; requesting them in the other formats responds with 404. They are shown with a dash in the download columns of their entries' pages.

Examples: /net/advogato/files/advogato.gt.zst, /net/iwdb/files/food_web_thompson_AkatoreA.csv.zip

Drawings

GET /draw/<name>[/<net>]

Returns a drawing of the network as a PNG image, or its thumbnail. The network name is only needed for entries with several networks. The drawings are made when the networks are processed; a request for a drawing that does not exist (yet) responds with 404.

Parameter	Value	Description	Default
`thumb`	boolean	Return the thumbnail (at most 1000×1000 pixels) instead of the full drawing.	false

Examples: /draw/iwdb/food_web_thompson_AkatoreA, /draw/iwdb/food_web_thompson_AkatoreA?thumb=True

Access from graph-tool

The graph-tool library wraps this API. Entry records are available as dictionaries:

import graph_tool.all as gt
gt.collection.ns_info["advogato"]

and networks are downloaded on demand. For entries with several networks, use a slash separator or a tuple:

g = gt.collection.ns["advogato"]
g = gt.collection.ns["iwdb/food_web_thompson_AkatoreA"]
g = gt.collection.ns[("iwdb", "food_web_thompson_AkatoreA")]