ts2net documentation
A Python implementation of time series to network methods for analyzing time series data through network science.
Overview
ts2net converts time series into networks using various methods:
Visibility Graphs: Horizontal Visibility Graph (HVG) and Natural Visibility Graph (NVG)
Recurrence Networks: Phase space embedding with recurrence analysis
Transition Networks: Symbolic dynamics and state transitions
Multivariate Networks: Networks from multiple time series using distance metrics
Ordinal Partition Networks: State space partitioning methods
Key Features
Fast implementations with Rust acceleration
Multiple network construction methods
Comprehensive visualization tools
Multivariate time series support
Windowed analysis for long time series
Integration with NetworkX for network analysis
Quick Start
import numpy as np
from ts2net import HVG, graph_summary
# Create a time series
x = np.sin(np.linspace(0, 12*np.pi, 800)) + 0.15 * np.random.randn(800)
# Build a Horizontal Visibility Graph
hvg = HVG()
hvg.build(x)
# Get network statistics
print(f"Nodes: {hvg.n_nodes}, Edges: {hvg.n_edges}")
G = hvg.as_networkx()
print(graph_summary(G))
Contents:
Guide