resid.etas: Residuals Analysis and Diagnostics Plots

In ETAS: Modeling Earthquake Data Using 'ETAS' Model

Residuals Analysis and Diagnostics Plots

Description

A function to compute and plot spatial and temporal residuals as well as transformed times for a fitted ETAS model.

Usage

   resid.etas(fit, type="raw", n.temp=1000, dimyx=NULL)

Arguments

fit	A fitted ETAS model. An object of class "etas".
type	A character string specifying the type residuals to be computed. Options are "raw" (the default case), "reciprocal" and "pearson".
n.temp	An integer specifying the number of partition points for temporal residuals.
dimyx	Dimensions of the discretization for the smoothed spatial residuals. A numeric vector of length 2.

Details

The function computes the temporal residuals

R^{temp}(I_j, h) = ∑_{i=1}^{N} δ_i 1[t_i \in I_j] h(t_i) λ^{temp}(t_i|H_{t_i}) - \int_{I_j} h(t)λ^{temp}(t|H_t) d t

for I_j=((j-1)T/n.temp, jT/n.temp], j=1,...,n.temp, and the (smoothed version of) spatial residuals

R^{spat}(B_j, h) = h(\tilde{x}_i, \tilde{y}_i) λ^{spat}(\tilde{x}_i, \tilde{y}_i)(\tilde{δ}_i - \tilde{w}_i)

for a Berman-Turner quadrature scheme with quadrature points (\tilde{x}_i, \tilde{y}_i) and quadrature weights \tilde{w}_i, i=1,...,n.spat. Raw, reciprocal and Pearson residuals obtain with h=1, h=1/λ and h=1/√{λ}, respectively.

In addition, the function computes transformed times

τ_j=\int_{0}^{t_j} λ^{temp}(t|H_t) d t

and

U_j = 1 - \exp(-(t_j - t_{j-1}))

Value

The function produces plots of temporal and smoothed spatial residuals, transformed times τ_j against j and Q-Q plot of U_j.

It also returns a list with components

• tau the transformed times

• U related quantities with the transformed times

• tres the temporal residuals

• sres the smoothed spatial residuals

Author(s)

Abdollah Jalilian [e-mail]

References

Baddeley A, Rubak E, Turner R (2015). Spatial Point Patterns: Methodology and Applications with R. Chapman and Hall/CRC Press, London. http://www.crcpress.com/Spatial-Point-Patterns-Methodology-and-Applications-with-R/Baddeley-Rubak-Turner/9781482210200/.

Baddeley A, Turner R (2000). Practical Maximum Pseudolikelihood for Spatial Point Patterns. Australian & New Zealand Journal of Statistics, 42(3), 283–322. doi: 10.1111/1467-842X.00128.

Baddeley A, Turner R, Moller J, Hazelton M (2005). Residual Analysis for Spatial Point Processes. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 67(5), 617–666. doi: 10.1111/j.1467-9868.2005.00519.x.

Ogata Y (1988). Statistical Models for Earthquake Occurrences and Residual Analysis for Point Processes. Journal of the American Statistical Association, 83(401), 9–27. doi: 10.2307/2288914.

Zhuang J (2006). Second-order Residual Analysis of Spatiotemporal Point Processes and Applications in Model Evaluation Journal of the Royal Statistical Society: Series B (Statistical Methodology), 68(4), 635–653. doi: 10.1111/j.1467-9868.2006.00559.x.

See Also

etas

Examples

  iran.cat <- catalog(iran.quakes, time.begin="1973/01/01",
     study.start="1986/01/01", study.end="2016/01/01",
     lat.range=c(26, 40), long.range=c(44, 63), mag.threshold=5)
  print(iran.cat)
  ## Not run: 
  plot(iran.cat)
## End(Not run)

  # setting initial parameter values
  param0 <- c(0.46, 0.23, 0.022, 2.8, 1.12, 0.012, 2.4, 0.35)

  # fitting the model
  ## Not run: 
  iran.fit <- etas(iran.cat, param0=param0)

  # diagnostic plots
  resid.etas(iran.fit)
## End(Not run)

ETAS documentation built on Nov. 28, 2022, 5:23 p.m.