changes {ssM8}R Documentation

Changes Made to the Package

Description

The M8 algorithm is based on seven different time series (see Keilis-Borok & Kossobokov, 1990; and Kossobokov, 1997). If the 7th time series and 5 out of the first 6 time series are above certain thresholds, a “Time of Increased Probability” (or TIP) is declared. This means there is an increased probability of an earthquake (of or greater than M0) in the next 5 years. Details for the calculation of the seven series can be found in under M8.series, and details for the calling of a TIP can be found under M8.TIP.

Li (1997) converted the original M8 FORTRAN code in Kossobokov (1997) into S-PLUS, which was then included into the Statistical Seismology Library (Harte, 1998). Ray Brownrigg subsequently packaged it into R.

Analyses using the M8 Algorithm, applied to New Zealand data, have been carried out by Ma & Vere-Jones (1997), Harte et al. (2003), and Harte et al. (2007).

This page contains a listing of recent changes made to functions, and known general problems.

Details

  1. decluster.M8: The line near the bottom within an else loop: class(minday) <- "datetimes" has been commented. (March 2002)
  2. plot.M8: Main title now uses default cex.main, vertical lines (type="h") have been removed, horizontal reference line changed to dashed (lty=2), colours removed from plotted symbols, and 8th graph plots only critical series. (January 2003)
  3. Occurrences of T and F have been changed to TRUE and FALSE, respectively. (January 2003)
  4. M8.TIP: When (training == "moving") the line TIP.type[1:7] <- NA has been changed to TIP.type[1:7] <- "". When (training == "user" || training == "all") the line TIP.type[1] <- NA has been changed to TIP.type[1] <- "". The earlier versions caused errors when running R CMD check in R version 1.6.2. (February 2003)
  5. plot.M8: argument M8 changed to x, and argument ... added to fix generic/method inconsistencies. (June 2003)
  6. print.mainshocks has been renamed to prt.mainshocks to fix generic/method confusion. (June 2003)
  7. time.grid has been renamed to timegrid to fix generic/method confusion. (June 2003)
  8. M8.TIP: the check to see if the critical series has length less than seven now uses function nrow rather than start and end. (October 2003)
  9. decluster.M8: a vector is now returned that identifies the mainshock to which each aftershock belongs. (October 2003)
  10. Minor formatting corrections to manual pages. (December 2003)
  11. Minor documentation formatting changes, mainly to use \dQuote. (January 2004)
  12. in.circle: correct documentation mismatch with code. (7 May 2004)
  13. decluster.M8: PACKAGE="ssM8" has been added to the .Fortran("recat", ...) call within the function. (7 May 2004)
  14. sum.row renamed to sumrow. (7 May 2004)
  15. A statement require(ssNZ) has been added to the examples that use the NZ catalogue. Hence ssNZ is no longer a “package” requirement of ssM8. (25 May 2004)
  16. M8.series: statement with incorrect logical syntax corrected: newline cover.breaks <- (time.breaks[1] <= catalogue[, "time"] <= time.breaks[n]). newline (17 Feb 2005)
  17. M8.TIP: make documentation consistent with Harte et al (2003). (17 Feb 2005)
  18. M8.TIP: method of calculating the quantiles in variable tops incorrect, redone using the function quantile. Values of g and h included in output list object. Variable initialisation rewritten. Code included to calculate TIP.level when training=="moving", previously not working for this option. (24 Mar 2005)
  19. emp.cdf: completely rewritten, see documentation. (11 Apr 2005)
  20. Old package ``stepfun'' not now ``required''. Part of base. (19 Apr 2005)
  21. decluster.M8: on manual page, units added for space and time windows. (9 Jul 2006)
  22. All occurences of subset.rect in the Examples have been changed to subsetrect. See Changes in ssBase. (8 Nov 2007)
  23. ssNZ added to Suggests in the DESCRIPTION file. Corresponding changes made to examples on manual pages. (10 Jan 2008)
  24. Use of require(ssNZ) in examples made consistent with package ssEDA. (28 Jan 2008)
  25. References updated below. (28 Jan 2008)
  26. Package vignettes added. (28 Jan 2008)

References

Harte, D. (1998). Documentation for the Statistical Seismology Library. School of Mathematical and Computing Sciences Research Report No. 98-10 (Updated Edition June 1999), Victoria University of Wellington. (ISSN 1174-4545)

Harte, D.; Li, D.-F.; Vreede, M. & Vere-Jones, D. (2003). Quantifying the M8 prediction algorithm: Reduction to a single critical variable and stability results. NZ Journal of Geology and Geophysics 46(1), 141–152. URL: http://www.rsnz.org/publish/nzjgg/2003/010.php

Harte, D.; Li, D.-F.; Vere-Jones, D.; Vreede, M. & Wang, Q. (2007). Quantifying the M8 algorithm: Model, forecast and evaluation. NZ Journal of Geology and Geophysics 50(2), 117–130. URL: http://www.rsnz.org/publish/nzjgg/2007/012.php

Keilis-Borok, V.I. and Kossobokov, V.G. (1990). Premonitory activation of earthquake flow: algorithm M8. Physics of the Earth and Planetary Interiors 61, 73–83. URL: http://dx.doi.org/10.1016/0031-9201(90)90096-G

Kossobokov, V.G. (1997). User manual for M8. In: Algorithms for Earthquake Statistics and Prediction. International Association of Seismology and Physics of the Earth's Interior (IASPEI) Library Volume 6. Pages 167–222. (Edited by J.H. Healy, V.I. Keilis-Borok and W.H.K. Lee). IASPEI, Menlo Park CA.

Li, Dongfeng. (1997). M8-Splus Documentation. Internal Report, Institute of Statistics and Operations Research, Victoria University of Wellington.

Ma, L. and Vere-Jones, D. (1997). Application of the M8 and Lin-Lin algorithms to New Zealand Earthquake data. NZ Journal of Geology and Geophysics 40, 77–89. URL: http://www.rsnz.org/publish/nzjgg/1997/7.php

[Package ssM8 version 2.1-3 Index]