Multivariate Granger causality and generalized variance

Physical Review E - Statistical, Nonlinear, and Soft Matter Physics

Volumn 81, Issue 4, 2010, Pages

  Barrett, Adam B ^a   Barnett, Lionel ^a   Seth, Anil K ^a  

^a UNIVERSITY OF SUSSEX   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

COMPLEX SYSTEMS; EXPERIMENTAL DATA; FUNCTIONAL RELATION; GENERALIZED VARIANCE; GRANGER CAUSALITY; RESIDUAL ERROR; SINGLE VARIABLE; UNIVARIATE;

COMPLEXATION;

EID: 77951117702     PISSN: 15393755     EISSN: 15502376     Source Type: Journal    
DOI: 10.1103/PhysRevE.81.041907     Document Type: Article

References (56)

1. M. Ding, Y. Chen, and S. Bressler, in Handbook of Time Series Analysis, edited by, S. Schelter, M. Winterhalder, and, J. Timmer, (Wiley, Wienheim, 2006), pp. 438-460.
2. K. Friston, L. Harrison, and W. Penny, Neuroimage 19, 1273 (2003). 10.1016/S1053-8119(03)00202-7
3. T. Schreiber, Phys. Rev. Lett. 85, 461 (2000). 10.1103/PhysRevLett.85.461
4. C. Ladroue, S. Guo, K. Kendrick, and J. Feng, PLoS ONE 4, e6899 (2009). 10.1371/journal.pone.0006899
5. J. Geweke, J. Am. Stat. Assoc. 77, 304 (1982). 10.2307/2287238
6. Z. Zhou, Y. Chen, M. Ding, P. Wright, Z. Lu, and Y. Liu, Hum. Brain Mapp 30, 2197 (2009). 10.1002/hbm.20661
7. N. Wiener, in Modern Mathematics for Engineers, edited by, E. F. Beckenbach, (McGraw Hill, New York, 1956).
8. C. W. J. Granger, Econometrica 37, 424 (1969). 10.2307/1912791
9. S. L. Bressler and A. K. Seth, Neuroimage (to be published).
10. A. K. Seth, J. Neurosci. Methods 186, 262 (2010). 10.1016/j.jneumeth. 2009.11.020
11. A. K. Seth, Cognitive Computation 1 (1), 50-63 (2009). 10.1007/s12559-009-9007-x
12. A. K. Seth, Network 16, 35 (2005). 10.1080/09548980500238756
13. A. K. Seth, Artif. Life 16, 179 (2009). 10.1162/artl.2010.16.2.16204
14. M. G. Kendall and A. Stuart, The Advanced Theory of Statistics (Griffin, London, 1979), Vol. 2.
15. This is to be distinguished from the conditional covariance, which will in general be a random variable, though later we note that for Gaussian variables the notions coincide.
16. L. Barnett, A. B. Barrett, and A. K. Seth, Phys. Rev. Lett. 103, 238701 (2009). 10.1103/PhysRevLett.103.238701
17. While our analysis may be extended to continuous time we focus here on the discrete-time case.
18. Here and in the remainder of this paper we assume, without loss of generality, that all random vectors and random processes have zero mean; thus constant terms are omitted in all linear regressions.
19. The analysis carries through for the nonstationary case, but for simplicity we assume here that all processes are stationary.
20. This might be more familiar as conditional G-causality, with Z as the conditioning variable. In practice it is the more useful form; for the nonconditional version, Z may simply be omitted.
21. - will generally be multivariate (at least if p, q > 1); hence they are written in bold type.
22. We remark that for significance testing of G-causality it is quite common to use the appropriate F statistic for regressions rather than F Y → X Z itself; the quantities are in any case related by a monotonic transformation.
23. C. W. J. Granger, Inf. Control. 6, 28 (1963). 10.1016/S0019-9958(63) 90092-5
24. P. Whittle, J. R. Stat. Soc. Ser. B (Methodol.) 15, 125 (1953).
25. A. Wald, Trans. Am. Math. Soc. 54, 426 (1943). 10.2307/1990256
26. J. Geweke, J. Am. Stat. Assoc. 79, 907 (1984). 10.2307/2288723
27. A. Kaiser and T. Schreiber, Physica D 166, 43 (2002). 10.1016/S0167-2789(02)00432-3
28. Y. Chen, G. Rangarajan, J. Feng, and M. Ding, Phys. Lett. A 324, 26 (2004). 10.1016/j.physleta.2004.02.032
29. D. Marinazzo, M. Pellicoro, and S. Stramaglia, Phys. Rev. Lett. 100, 144103 (2008). 10.1103/PhysRevLett.100.144103
30. If the predictee variable has a continuous (multivariate) distribution, we note that the Jacobian determinants in the standard change-of-variables formula for entropy calculation cancel out.
31. O. Sporns, G. Tononi, and G. M. Edelman, Cereb. Cortex 10, 127 (2000). 10.1093/cercor/10.2.127
32. Y. Chen, S. L. Bressler, and M. Ding, J. Neurosci. Methods 150, 228 (2006). 10.1016/j.jneumeth.2005.06.011
33. Y. A. Rozanov, Stationary Random Processes (Holden-Day, San Francisco, CA, 1967).
34. J. Doob, Stochastic Processes (Wiley, New York, 1953).
35. -) which appears in F Y → X (using a finite number of lags will generally result in an overestimate of F Y → X since residual errors will be larger than for the exact regression). As applied to empirical data, it is in any case good practice to choose "sufficient" lags for all regressions so as to model the data adequately without overfitting.
36. S. Guo, A. K. Seth, K. Kendrick, C. Zhou, and J. Feng, J. Neurosci. Methods 172, 79 (2008). 10.1016/j.jneumeth.2008.04.011
37. L. A. Baccalá and K. Sameshima, Biol. Cybern. 84, 463 (2001). 10.1007/PL00007990
38. In it is stated that partial G-causality may in some circumstances be negative; the justification for this is unclear.
39. In, under the section headed "partial complex Granger causality," the quantity developed appears to be (the trace version of) what is conventionally referred to as conditional G-causality rather than partial G-causality as introduced in and referenced in this section.
40. This is the "weighted" version of causal density. An unweighted and [0,1] bounded alternative can be defined as the fraction of all pairwise conditional causalities that are statistically significant at a given significance level.
41. O. Sporns, Scholarpedia J. 2, 1623 (2007). 10.4249/scholarpedia.1623
42. A. K. Seth, E. Izhikevich, G. N. Reeke, and G. M. Edelman, Proc. Natl. Acad. Sci. U.S.A. 103, 10799 (2006). 10.1073/pnas.0604347103
43. M. Shanahan, Phys. Rev. E 78, 041924 (2008). 10.1103/PhysRevE.78.041924
44. G. Tononi, O. Sporns, and G. M. Edelman, Proc. Natl. Acad. Sci. U.S.A. 91, 5033 (1994). 10.1073/pnas.91.11.5033
45. M. A. Bedau, in Philosophical Perspectives: Mind, Causation, and World, edited by, James Tomberlin, (Oxford: Blackwell Publishers, Oxford, UK, 1997), 11, 375.
46. C. Shalizi and C. Moore, e-print arXiv:cond-mat/0303625.
47. C. R. Shalizi, R. Haslinger, J.-B. Rouquier, K. L. Klinkner, and C. Moore, Phys. Rev. E 73, 036104 (2006). 10.1103/PhysRevE.73.036104
48. A. K. Seth, Scholarpedia J. 2, 1667 (2007). 10.4249/scholarpedia.1667
49. D. O. Hebb, The Organization of Behavior (Wiley, New York, 1949).
50. K. D. Harris, Nat. Rev. Neurosci. 6, 399 (2005). 10.1038/nrn1669
51. In approximation.
52. R. Gaillard, S. Dehaene, C. Adam, S. Clémenceau, D. Hasboun, M. Baulac, L. Cohen, and L. Naccache, PLoS Biol. 7, e61 (2009). 10.1371/journal.pbio.1000061
53. L. Angelini, M. de Tommaso, D. Marinazzo, L. Nitti, M. Pellicoro, and S. Stramaglia, Phys. Rev. E (to be published).
54. R. B. Kline, Principles and Practice of Structural Equation Modeling (Guilford Press, New York, 2005).
55. H. Akaike, IEEE Trans. Autom. Control 19, 716 (1974). 10.1109/TAC.1974.1100705
56. G. Schwarz, Ann. Stat. 6, 461 (1978). 10.1214/aos/1176344136