This toolkit provides functions for building, transforming, and analyzing temporal networks.
It is organized into five functional groups that form a pipeline:

Convert dynamic functional connectivity (DFC) into temporal networks.

• ComputeDynamicFunctionalConnectivity(time_series, window, lag)
  → Sliding-window Pearson correlation matrices.
• FindThresholdForAverageDegree(adj_tensor, target_degree, resolution=100)
  → Find threshold to match a target average degree.
• BinarizeDFC(time_series, target_degree, window, lag)
  → Pipeline: DFC → thresholding → binary adjacency.
• SetDiagonals(tensor, value=1)
  → Utility to enforce diagonal values.

Create synthetic and null temporal networks.

• GenerateSymmetricRandomNetwork
• GenerateSymmetricSmallWorldNetwork
• GenerateSymmetricScaleFreeNetwork
• GenerateSymmetricNetwork (dispatch by type)
• GenerateStaticTemporalNetwork
• GenerateTemporalNetworkByLinkActivation
• GenerateTemporalNetworkWithDensityVariability
• GenerateTemporalLinkCounts
• GenerateNullModel
• PartiallyRandomizeMatrix
• GenerateRandomizedTemporalNetwork
• TrimIsolatedNodes

Temporal reachability, shortest paths, and circulation.

• ComputeAverageDegree
• SmartWalker → Earliest arrival latencies.
• RandomWalker → Monte Carlo first-passage latencies.
• MeanLatencyMatrixAnalysis → Summarize latency matrix.
• ComputeTemporalDistanceMeasures → Bundle (degree + walkers).
• ComputeCirculationLatency
• ComputeCirculationRate
• ComputeCirculationLatencyAndRate (efficient one-pass).

Quantify network volatility and temporal memory.

• ComputeTemporalMutualInformation
• ComputeAdjustedEntropy
• ComputeTemporalEdgeOverlap
• ComputeEdgePersistenceRate
• ComputeNeighborhoodMemory
• ComputeReturnability
• ComputeLinkBurstiness

Measure clustering, modularity, and persistence of community structure.

• ComputeStaticClustering
• ComputeTemporalClustering
• ComputeSnapshotTransitivity, ComputeTemporalTransitivity
• ComputeSnapshotParticipationCoefficient, ComputeTemporalParticipationCoefficient
• ComputePartnerStability
• ComputePartnerDiversity
• ComputeNodePersistence

• ComputeSnapshotModularity, ComputeTemporalModularity

1. Time series → TNet (Group 1)
2. Null models & synthetic networks (Group 2)
3. Distance & circulation metrics (Group 3)
4. Dynamism & memory analysis (Group 4)
5. Segregation & cohesion analysis (Group 5)

Together these provide a unified framework for studying both integration and segregation in temporal networks.

Built by Simachew Mengiste (with Demian Bataglia)

Driven by curiosity, scientific clarity, and modular design. FunSy - LNCA - Unistra