linksrm_gif(data, evalpts, params, tplus=FALSE, eta=0.75)
linksrm_gif(data, evalpts=NULL, params, TT, eta=0.75)linksrm1_gif(data, evalpts, params, tplus=FALSE, eta=0.75)
linksrm1_gif(data, evalpts=NULL, params, TT, eta=0.75)
"time", usually the number of days from some origin; "magnitude" which is the event magnitude less the magnitude thresholmatrix or data.frame. It must include two columns named "time" and "region" that can be referred to as evalpts[,"time"]TRUE, else $\lambda_g(t^-,i|{\cal H}_t)$.The difference between linksrm_gif and linksrm1_gif is that the stress reduction matrices St1 and St2 (internal to both functions) are calculated every time that the function is called in the case of linksrm_gif. If the event history is not changing between successive calls (e.g. parameter estimation), then this is unnecessary. However, in a simulation, the history changes with the addition of each new event. The function linksrm1_gif checks to see whether the matrices St1 and St2 exist. If so, these existing matrices are used, and new ones are not calculated. Therefore when using linksrm1_gif for parameter estimation, one must check for the existence of such matrices, and delete upon starting to fit a new model, for example: if (exists("St1")) rm(St1)
if (exists("St2")) rm(St2)
St1 and St2 could be dealt with in a tidier manner. This is the only difference between linksrm_gif and linksrm1_gif.
}
gif.