Fig. 1
From: Using bundle embeddings to predict daily cortisol levels in human subjects
Construction of an embedding. In this example, with e = 3 and τ = 1, three-dimensional coordinate vectors are produced from time series x(t). The scalar components of each vector are obtained by taking lagged values of x at time t, t-1 and t-2. The embedding E is the set of all generated coordinate vectors. A dynamical path, connecting subsequent vectors in the embedding, has been omitted for the sake of clarity