Skip to contents

Recurrence quantification analysis (RQA)

Source: R/stability-orbital-rqa.R
recurrenceQuantification.Rd

Reconstructs the attractor by delay embedding and quantifies its recurrence structure: recurrence rate, determinism, laminarity and diagonal/vertical line statistics. High determinism indicates deterministic (periodic/chaotic) rather than stochastic dynamics.

Usage

recurrenceQuantification(
  x,
  m = 3L,
  tau = 1L,
  radius = NULL,
  target_rr = 0.1,
  lmin = 2L,
  vmin = 2L,
  max_n = 1500L
)

Arguments

x

Numeric time series.

m

Embedding dimension (default 3).

tau

Embedding delay (default 1).

radius

Recurrence threshold; if NULL, chosen to reach target_rr.

target_rr

Target recurrence rate when radius is NULL (default 0.1).

lmin, vmin

Minimum diagonal / vertical line lengths (default 2).

max_n

Cap on embedded points (subsampled if longer; default 1500).

Value

an rqa_result: RR, DET, LAM, Lmax, Lmean, TT (trapping time), ENTR (diagonal line entropy), radius.

References

Marwan N, et al. (2007) Phys Rep 438:237-329.

See also

floquetMultipliers(), approximateEntropy()

Examples

t <- seq(0, 20 * pi, length.out = 800)
recurrenceQuantification(sin(t))$DET               # ~1 (deterministic)
#> [1] 0.9536178