Surrogate Data with Correlations, Trends, and Nonstationarities

The data in this collection include: (1) 6 surrogate stationary signals with different correlations; (2) 7 surrogate correlated signals with linear, sinusoidal and power-law trends; and (3) 15 surrogate correlated signals with different types of nonstationarities. Each data file contains one column of data in ASCII format. Results on correlated signals with trends are discussed in Physical Review E 64, 011114 (2001). Results on correlated signals with different types of nonstationarities are discussed in Physical Review E 65, 041107 (2002). The parameter "alpha" (see below) is an exponent measuring the degree of correlations in a signal, and Nmax is the signal length. A detailed description of these signals can be found in the original articles.

Correlations in these signals can be quantified using Detrended Fluctuation Analysis (DFA). Limitations of the DFA method are discussed in the articles cited above. In particular, the second paper notes that

... for anti-correlated signals, the scaling exponent obtained from the DFA method overestimates the true correlations at small scales. To avoid this problem, one needs first to integrate the original anti-correlated signal and then apply the DFA method. The correct scaling exponent can thus be obtained from the relation between n [the DFA box length] and F(n)/n instead of F(n) ... In order to provide a more accurate estimate of F(n), the largest box size n we use is N_max/10, where N_max is the total number of points in the signal.

Since these files are quite large, they are provided as gzip-compressed text.

1. Correlated stationary signals

• noise0117.txt.gz     alpha = 0.1, N_max = 2¹⁷;
• noise0217.txt.gz     alpha = 0.2, N_max = 2¹⁷;
• noise0517.txt.gz     alpha = 0.5, N_max = 2¹⁷;
• noise0817.txt.gz     alpha = 0.8, N_max = 2¹⁷;
• noise0917.txt.gz     alpha = 0.9, N_max = 2¹⁷;
• noise1517.txt.gz     alpha = 1.5, N_max = 2¹⁷.

2. Surrogate signals with trends

    2a) Signals with linear trends

• trlina1.txt.gz     alpha = 0.1, N_max = 2¹⁷, slope of linear trend A_l = 2^-16 / index;
• trlina2.txt.gz     alpha = 0.1, N_max = 2¹⁷, slope of linear trend A_l = 2^-12 / index;
• trlina3.txt.gz     alpha = 0.1, N_max = 2¹⁷, slope of linear trend A_l = 2^-8 / index.

• trsin1.txt.gz     alpha = 0.9, N_max = 2¹⁷, Amplitude of trend A_s = 2, period T = 128;
• trsin2.txt.gz     alpha = 0.1, N_max = 2¹⁷, Amplitude of trend A_s = 2, period T = 128.

• trpow1.txt.gz     alpha = 0.9, N_max = 2¹⁷, power lambda = 0.4, Amplitude A_p = 1000 / (N_max) ^lambda;
• trpow2.txt.gz     alpha = 1.5, N_max = 2¹⁷, power lambda = -0.7, Amplitude A_p = 0.01 / (N_max) ^lambda.

3. Surrogate nonstationary signals

    3a) Signals with cutout segments (discontinuities)

• cut0117w20p95.txt.gz     alpha = 0.1, seg. cutout probability p = 0.05, Width W = 20, N_max = 2¹⁷;
• cut0117w20p50.txt.gz     alpha = 0.1, seg. cutout probability p = 0.50, Width W = 20, N_max = 2¹⁷;
• cut0917w20p95.txt.gz     alpha = 0.9, seg. cutout probability p = 0.05, Width W = 20, N_max = 2¹⁷;
• cut0917w20p50.txt.gz     alpha = 0.9, seg. cutout probability p = 0.50, Width W = 20, N_max = 2¹⁷.

• sp02p05a1.txt.gz           spikes probability p = 0.05, Amplitude Asp = 1, N_max = 2¹⁷;
• sp02p05a1sp.txt.gz       spikes signal only, spikes probability p = 0.05, Amplitude Asp = 1, N_max = 2¹⁷;
• sp08p05a10.txt.gz         spikes probability p = 0.05, Amplitude Asp = 10, N_max = 2¹⁷;
• sp08p05a10sp.txt.gz     spikes signal only, spikes probability p = 0.05, Amplitude Asp = 10, N_max = 2¹⁷.

• d2h4pd050118s.txt.gz     alpha = 0.1, sigma₁ = 1, sigma₂ = 4 (probability p = 0.05), N_max = 2¹⁸;
• d2h4pd950118s.txt.gz     alpha = 0.1, sigma₁ = 1, sigma₂ = 4 (probability p = 0.95), N_max = 2¹⁸;
• d2h4pd050918s.txt.gz     alpha = 0.9, sigma₁ = 1, sigma₂ = 4 (probability p = 0.05), N_max = 2¹⁸;
• d2h4pd950918s.txt.gz     alpha = 0.9, sigma₁ = 1, sigma₂ = 4 (probability p = 0.95), N_max = 2¹⁸.

• cut010917p90w20_sum.txt.gz (mixed signal)          alpha₁ = 0.1 (90%), alpha₂ = 0.9(10%), Width = 20, N_max = 2¹⁷;
• cut010917p90w20_comp1.txt.gz (component 1)     alpha₁ = 0.1 (90%) only, Width W = 20, N_max = 2¹⁷;
• cut010917p90w20_comp2.txt.gz (component 2)     alpha₂ = 0.9 (10%) only, Width W = 20, N_max = 2¹⁷.

 
Address for correspondence:
Plamen Ch. Ivanov, Ph.D.
Room 247, Dept. of Physics
Boston Univeristy
590 Commonwealth Avenue
Boston, MA 02215, USA
Email: plamen@meta.bu.edu

Your comments and suggestions are welcome. We encourage you to use our feedback form to comment on this page. If you would like to receive a reply, please send your comments by email to webmaster@physionet.org, or post them to: PhysioNet MIT Room E25-505A 77 Massachusetts Avenue Cambridge, MA 02139 USA

Updated Friday, 07 March 2003 at 14:19 EST