# NOT RUN {
data(trentino)
year_min <- 1961
year_max <- 1990
period <- PRECIPITATION$year>=year_min & PRECIPITATION$year<=year_max
station <- names(PRECIPITATION)[!(names(PRECIPITATION) %in% c("day","month","year"))]
prec_mes <- PRECIPITATION[period,station]
## removing nonworking stations (e.g. time series with NA)
accepted <- array(TRUE,length(names(prec_mes)))
names(accepted) <- names(prec_mes)
for (it in names(prec_mes)) {
accepted[it] <- (length(which(!is.na(prec_mes[,it])))==length(prec_mes[,it]))
}
prec_mes <- prec_mes[,accepted]
## the dateset is reduced!!!
prec_mes <- prec_mes[,1:2]
p <- 1 ## try p <- 2 !!!
CCGamma <- CCGammaToBlockmatrix(data=prec_mes,lag=0,p=p,tolerance=0.001)
# }
# NOT RUN {
## Not Run in the examples, uncomment to run the following line
CCGamma_1 <- CCGammaToBlockmatrix(data=prec_mes,lag=1,p=p,tolerance=0.001)
### Alternatively, recommended .....
## Not Run in the examples, uncomment to run the following line
CCGamma <- CCGammaToBlockmatrix(data=prec_mes,lag=0,p=p+1,tolerance=0.001)
CCGamma0 <- CCGamma[1:p,1:p]
CCGamma1 <- CCGamma[(1:p),(1:p)+1]
CCGamma0_inv <- solve(CCGamma0)
## Not Run in the examples, uncomment to run the following line
a1 <- blockmatmult(CCGamma0,CCGamma0_inv)
a2 <- blockmatmult(CCGamma1,CCGamma0_inv)
CCGamma_1t <- t(CCGamma1)
CCGamma_0t <- t(CCGamma0)
A <- t(solve(CCGamma_0t,CCGamma_1t))
# }
# NOT RUN {
# }
Run the code above in your browser using DataLab