NA_LM Model

# Cut-and-paste code below into window above and Run

#

# North America (NA) Model (1950-2000)

#

# Measurement Matrix    (Growth), (N-XREAL-X-L), (XREAL-X-L+N)

#

#           Q     N  XREAL      X      L

#[1,]  0.4544 0.443  0.441  0.445  0.453

#[2,]  0.2183 0.555 -0.574 -0.487  0.276

#[3,] -0.0995 0.412  0.623 -0.530 -0.389

#

# Fraction of Variance 

#[1] 0.960 0.996 1.000 1.000 1.000

#

merge.forecast <- function (fx,n=1) {

x <- splice(fx$pred,fx$forecast[[n]])

colnames(x) <- seriesNames(fx$data$output)

return(x)

}

AIC <- function(model) {informationTestsCalculations(model)[3]}

require(dse)

require(matlab)

f <- matrix( c(1.034221555, -0.00410808, -0.02808741,  0.16057008,

                     -0.033299694,  1.05703398,  0.07032517, -0.01391041,

                      0.003681037, -0.06723139,  0.99583196,  0.01197887,

0.00000000,  0.0000000,  0.0000000,  1.0000000000

),byrow=TRUE,nrow=4,ncol=4)

#

# To stabilize model, uncomment next line

# f[1,1] <- 0.993469099; f[2,2] <- 1.01538263; f[3,3] <- 0.95659221

#

h <- eye(3,4)

k <- (f[,1:3,drop=FALSE])

NA_LM <- SS(F=f,H=h,K=k,z0=c(0.16057008, -0.01391041,  0.01197887,  1.0000000000),

              output.names=c("NA1","NA2","NA3"))

print(NA_LM)

is.SS(NA_LM)

stability(NA_LM)

NA_LM.data <- simulate(NA_LM,sampleT=100,start=1950)

#NA_LM.data <- simulate(NA_LM,sampleT=20,noise=matrix(0,20,3))

NA_LM.f <- forecast(l(NA_LM,NA_LM.data),horizon=150)

MA_LM.fx <- merge.forecast(NA_LM.f)

tfplot(NA_LM.f)

AIC(m <-  l(NA_LM,NA_LM.data))

NA_LMx <- SS(F=f,H=h,Q=eye(4,3),R=eye(3,3),z0=c(0.16057008, -0.01391041,  0.01197887,  1.0000000000),

              output.names=c("NA1","NA2","NA3"))

     shockDecomposition(NA_LMx)