Guide
annnet keeps graph structure, annotations, and analysis contexts in one object. These five notebooks explain how those parts work together. Each topic includes the concepts, relevant mathematics, working examples, and links to the API.
Start with the quickstart, then read the topics you need:
| Topic | What you will learn |
|---|---|
| Graph model and matrices | Node and edge identity, hyperedges, incidence coefficients, adjacency, and traversal. |
| Annotations, selections, and views | Where attributes live, table access, filtering, live views, and materialized graphs. |
| Slices and analysis contexts | Keep a prior network and condition-specific selections and results together. |
| Layers | Aspects and placements, importing a multilayer graph, measurements, coupling, supra-matrices, and dynamics. |
| Computation and storage | Run backend algorithms, attach results, choose an export, save an analysis, and understand history limits. |
Slices select existing identities; layers place identities in explicit coordinates. Use slices when two analyses select or annotate the same interactions. Use layers when an entity's placement in a condition, mechanism, or timepoint is itself part of the graph. The two can be used together.
Run the notebooks
Install the package with the dependencies used by the guide:
Alternatively, use the conda environment. Download a notebook
from its page, and put the three sample CSV files in a
data/ directory beside it. In a repository checkout, the files are already in
docs/guide/data. Run the cells in order; each notebook imports its own graph
and does not depend on the other notebooks.
The data are small and synthetic, so these notebooks require no external data service. For longer analyses with biological data or other software, see Tutorials.