Nothing
R/rmgarch-mmean.R
In rmgarch: Multivariate GARCH Models
Defines functions
mvmean.arsim
mvmean.varsim
mvmean.arforecast
mvmean.varforecast
mvmean.arfilter
mvmean.varfilter
mvmean.arfit
mvmean.varfit
#################################################################################
##
## R package rmgarch by Alexios Galanos Copyright (C) 2008-2022.
## This file is part of the R package rmgarch.
##
## The R package rmgarch is free software: you can redistribute it and/or modify
## it under the terms of the GNU General Public License as published by
## the Free Software Foundation, either version 3 of the License, or
## (at your option) any later version.
##
## The R package rmgarch is distributed in the hope that it will be useful,
## but WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
## GNU General Public License for more details.
##
#################################################################################
# First Stage Conditional Mean Filter
# ggn = goggarch numbering (+2 over copula and dcc models)
mvmean.varfit = function(model, data, VAR.fit = NULL, T, out.sample = 0, cluster = NULL, ggn = 0)
{
if( model$modelinc[2+ggn]>0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] ) stop("\ngogarchfit-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
if( !is.null(VAR.fit) ){
zdata = VAR.fit$xresiduals
# should be N - p - n.start
model$modelinc[1+ggn] = p = mp = VAR.fit$lag
N = dim(zdata)[1]
if(N != T) stop(paste("\ngogarchfit-->error: Residuals of VAR.fit object should be of length ", T, " (N - out.sample) ", sep=""))
mu = VAR.fit$xfitted
varcoef = VAR.fit$Bcoef
} else{
ic = model$varmodel$lag.criterion[1]
if( !is.null(model$varmodel$lag.max) ){
mp = .varxselect(y = data[1:T, ], lag.max = model$varmodel$lag.max, exogen = ex[1:T, , drop=FALSE])$selection
mp = as.numeric( mp[which(substr(names(mp), 1, 2) == substr(ic, 1, 2))] )
} else {
mp = model$modelinc[1+ggn]
}
p = mp
varmodel = varxfit(X = data[1:T, ,drop=FALSE], p = mp, constant = TRUE, exogen = ex[1:T, , drop=FALSE],
robust = model$varmodel$robust, gamma = model$varmodel$robust.control$gamma,
delta = model$varmodel$robust.control$delta, nc = model$varmodel$robust.control$nc,
ns = model$varmodel$robust.control$ns, postpad = "constant", cluster = cluster)
zdata = varmodel$xresiduals
# should be T
N = dim(zdata)[1]
mu = varmodel$xfitted
model$modelinc[1+ggn] = mp
varcoef = varmodel$Bcoef
}
# Filter Full Dataset now
if(out.sample > 0){
varmodel2 = varxfilter(X = data, p = mp, exogen = ex, Bcoef = varcoef, postpad = "constant")
zdata = varmodel2$xresiduals
# should be T - p + out.sample
N = dim(zdata)[1]
mu = varmodel2$xfitted
} else{
zdata = zdata
}
model$varcoef = varcoef
return(list(model = model, zdata = zdata, mu = mu, varcoef = varcoef, p = p, N = N))
}
# arfit not used by dcc and copula models
mvmean.arfit = function(model, data, ARcoef = NULL, T, out.sample = 0, solver,
solver.control = list(), fit.control = list(), cluster = NULL)
{
m = NCOL(data)
if(is.null(ARcoef)){
if( model$modelinc[4] > 0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfit-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
p = model$modelinc[2]
# Irrespsective of the Distribution, the mean filtration is always
# Normal.
arspec = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex[1:T, , drop = FALSE]),
distribution.model = "norm")
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("T", "arspec", "data", "solver",
"solver.control", "fit.control"), envir = environment())
arfit = parLapply(cluster, as.list(1:m), fun = function(i){
rugarch::arfimafit(spec = arspec, data = data[1:T,i,drop=FALSE],
out.sample = 0, solver = solver,
solver.control = solver.control,
fit.control = fit.control)
})
} else{
arfit = lapply(1:m, FUN = function(i) arfimafit(spec = arspec,
data = data[1:T,i,drop=FALSE], out.sample = 0,
solver = solver, solver.control = solver.control,
fit.control = fit.control))
}
zdata = sapply(arfit, FUN = function(x) residuals(x))
#cat("done!\n")
# rugarch always returns full length data
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
# should be T - p
N = NROW(zdata)
mu = sapply(arfit, FUN = function(x) fitted(x))
#if( p > 0 ) mu = mu[-c(1:p), , drop = FALSE]
arcoef = sapply(arfit, FUN = function(x) coef(x))
} else{
if( model$modelinc[4] > 0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfit-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
p = model$modelinc[2]
# Irrespsective of the Distribution, the mean filtration is always Normal.
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("T", "ARcoef", "data", "solver",
"solver.control", "fit.control"), envir = environment())
arfit = parLapply(cluster, as.list(1:m), fun = function(i){
arspec = rugarch::arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex[1:T, , drop = FALSE]),
distribution.model = "norm", fixed.pars = as.list(ARcoef[,i]));
ans = rugarch::arfimafilter(spec = arspec, data = data[1:T,i,drop=FALSE],
out.sample = 0);
return(ans)
})
} else{
arfit = lapply(1:m, FUN = function(i){
arspec = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex[1:T, , drop = FALSE]),
distribution.model = "norm", fixed.pars = as.list(ARcoef[,i]));
ans = arfimafilter(spec = arspec, data = data[1:T,i,drop=FALSE], out.sample = 0);
return(ans)
})
}
zdata = sapply(arfit, FUN = function(x) residuals(x))
#cat("done!\n")
# rugarch always returns full length data
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
# should be T - p
N = NROW(zdata)
mu = sapply(arfit, FUN = function(x) fitted(x))
#if( p > 0 ) mu = mu[-c(1:p), , drop = FALSE]
arcoef = ARcoef
}
if(out.sample>0){
if( !is.null(cluster) ){
clusterExport(cluster, c("data", "p", "ex", "arcoef"),
envir = environment())
arfit2 = parLapply(cluster, as.list(1:m), fun = function(i) {
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]));
ans = arfimafilter(spec = specx, data = data[,i, drop=FALSE], out.sample = 0);
return(ans)
})
} else{
arfit2 = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex), distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]));
ans = arfimafilter(spec = specx, data = data[,i, drop = FALSE], out.sample = 0);
return(ans)
})
}
zdata = sapply(arfit2, FUN = function(x) residuals(x))
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
N = NROW(zdata)
} else{
zdata = zdata
}
model$arcoef = arcoef
return(list(model = model, zdata = zdata, mu = mu, arcoef = arcoef, p = p, N = N))
}
mvmean.varfilter = function(model, data, varcoef, T, out.sample = 0, ggn = 0)
{
if( model$modelinc[2+ggn]>0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfilter-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
if(is.null(varcoef))
stop("\ncgarchfilter-->error: VAR model requires varcoef in specification to be supplied.")
p = model$modelinc[1+ggn]
vrmodel = varxfilter(X = data[1:T, , drop=FALSE], p = p, exogen = ex[1:T, , drop=FALSE], Bcoef = varcoef, postpad = "constant")
#cat("done!\n")
zdata = vrmodel$xresiduals
# should be N - p
N = NROW(data)
mu = vrmodel$xfitted
# if we have out of sample data we need to filter
if(out.sample > 0){
varmodel2 = varxfilter(X = data, p = p, exogen = ex, Bcoef = varcoef, postpad = "constant")
zdata = varmodel2$xresiduals
N = NROW(zdata)
mu = varmodel2$xfitted
} else{
zdata = zdata
}
model$varcoef = varcoef
model$residuals = zdata
return(list(model = model, zdata = zdata, mu = mu, p = p, N = N))
}
mvmean.arfilter = function(model, data, arcoef, T, out.sample = 0, cluster = NULL)
{
if( model$modelinc[4] > 0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfilter-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
m = NCOL(data)
p = model$modelinc[2]
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("T", "data", "p", "ex", "arcoef"),
envir = environment())
armodel = parLapply(cluster, as.list(1:m), fun = function(i) {
specx = rugarch::arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex[1:T, ]),
distribution.model = "norm",
fixed.pars = as.list(arcoef[, i]))
ans = rugarch::arfimafilter(spec = specx, data = data[1:T, i], out.sample = 0)
return(ans)
})
} else{
armodel = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex[1:T, ]), distribution.model = "norm",
fixed.pars = as.list(arcoef[, i]))
ans = arfimafilter(spec = specx, data = data[1:T, i], out.sample = 0)
return(ans)
})
}
#cat("done!\n")
zdata = sapply(armodel, FUN = function(x) residuals(x))
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
# should be N - p
N = NROW(data)
mu = sapply(armodel, FUN = function(x) fitted(x))
#if( p > 0 ) mu = mu[-c(1:p), , drop = FALSE]
# if we have out of sample data we need to filter
if(out.sample > 0){
if( !is.null(cluster) ){
clusterExport(cluster, c("data", "p", "ex", "arcoef"),
envir = environment())
armodel2 = parLapply(cluster, as.list(1:m), fun = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimafilter(spec = specx, data = data[,i], out.sample = 0)
return(ans)
})
} else{
armodel2 = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimafilter(spec = specx, data = data[,i], out.sample = 0)
return(ans)
})
}
zdata = sapply(armodel2, FUN = function(x) residuals(x))
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
} else{
zdata = zdata
}
return(list(zdata = zdata, mu = mu, p = p, N = N))
}
mvmean.varforecast = function(model, mregfor, varcoef, n.ahead, n.roll, ns, ggn = 0){
m = NCOL(model$modeldata$data)
tf = n.ahead + n.roll
if( model$modelinc[2+ggn] > 0 ){
if( is.null( mregfor ) ){
warning("\nExternal Regressor Forecasts Matrix NULL...setting to zero...\n")
mregfor = matrix(0, ncol = model$modelinc[2+ggn], nrow = (n.roll + n.ahead) )
} else{
if( dim(mregfor)[2] != model$modelinc[2+ggn] )
stop("\ngogarchforecast-->error: wrong number of external regressors!...", call. = FALSE)
if( dim(mregfor)[1] < (n.roll + n.ahead) )
stop("\ngogarchforecast-->error: external regressor matrix has less points than requested forecast length (1+n.roll) x n.ahead!...", call. = FALSE)
}
} else{
mregfor = NULL
}
if( n.ahead == 1 && (n.roll > ns) )
stop("\ngogarchforecast-->error: n.roll greater than out.sample!", call. = FALSE)
# varxforecast returns array --> change to matrix since we do not allow
# mixed n.roll and n.ahead
Mu = varxforecast(X = model$modeldata$data, Bcoef = varcoef, p = model$modelinc[1+ggn], out.sample = ns, n.ahead, n.roll, mregfor)
return(Mu)
}
mvmean.arforecast = function(model, mregfor, arcoef, n.ahead, n.roll, ns, cluster = NULL)
{
m = NCOL(model$modeldata$data)
tf = n.ahead + n.roll
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("model", "arcoef", "mregfor", "ns",
"n.ahead", "n.roll"), envir = environment())
arfx = parLapply(cluster, as.list(1:m), fun = function(i){
specx = rugarch::arfimaspec(mean.model = list(armaOrder = c(model$modelinc[2], 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = model$modeldata$mexdata),
distribution.model = "norm", fixed.pars = as.list(arcoef[,i]))
ans = rugarch::arfimaforecast(fitORspec = specx, data = model$modeldata$data[,i],
out.sample = ns, n.ahead = n.ahead, n.roll = n.roll,
external.forecasts = list(mregfor = mregfor))
return(ans)
})
} else{
arfx = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(model$modelinc[2], 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = model$modeldata$mexdata),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimaforecast(fitORspec = specx, data = model$modeldata$data[,i],
out.sample = ns, n.ahead = n.ahead, n.roll = n.roll,
external.forecasts = list(mregfor = mregfor))
return(ans)
})
}
Mu = array(NA, dim=c(n.ahead, m, n.roll+1))
for(i in 1:(n.roll+1)){
Mu[,,i] = matrix(sapply(arfx, FUN = function(x) fitted(x)[,i], simplify=TRUE), ncol = m)
}
return(Mu)
}
mvmean.varsim = function(model, Data, res, mexsimdata, prereturns, m.sim,
n.sim, n.start, startMethod = "sample", cluster = NULL, ggn = 0)
{
seriesSim = vector(mode = "list", length = m.sim)
m = NCOL(Data)
varcoef = model$varcoef
if( model$modelinc[2+ggn] > 0 ){
if( !is.null(mexsimdata) ){
if( !is.list(mexsimdata) | length(mexsimdata) != m.sim )
stop("\ngogarchsim-->error: mexsimdata must be a list of length equal to m.sim...", call. = FALSE)
for(j in 1:m.sim){
if( !is.matrix(mexsimdata[[j]]) )
stop("\ngogarchsim-->error: mexsimdata list submatrix must be a matrix...", call. = FALSE)
if( dim(mexsimdata[[j]])[2] != model$modelinc[2+ggn] )
stop("\ngogarchsim-->error: wrong number of external regressors in list submatrix!...", call. = FALSE)
if( dim(mexsimdata[[j]])[1] < (n.sim + n.start) )
stop("\ngogarchsim-->error: external regressor list submatrix has less points than requested simulation length (n.sim + n.start)...", call. = FALSE)
}
} else{
mexsimdata = vector(mode = "list", length = m.sim)
for(j in 1:m.sim) mexsimdata[[j]] = matrix(0, ncol = model$modelinc[2+ggn], nrow = (n.sim + n.start))
}
} else{
mexsimdata = NULL
}
p = model$modelinc[1+ggn]
if(startMethod == "sample"){
if( !is.null(prereturns) ){
if(!is.matrix(prereturns))
stop("\nprereturns must be a matrix\n")
if(dim(prereturns)[1] < p)
stop("\nwrong number of rows for prereturns matrix")
if(dim(prereturns)[2] != m)
stop("\nwrong number of cols for prereturns matrix")
prereturns = matrix(tail(prereturns, p), nrow = p, ncol = m)
} else{
prereturns = as.matrix( tail(Data, p) )
}
} else{
prereturns = matrix(0, nrow = p, ncol = m, byrow = TRUE)
}
if( n.sim == 1 && n.start == 0 ){
# SPECIAL ROUTINE TO SPEED UP 1-AHEAD ESTIMATION
tmp = varxsimXX(X = Data, varcoef, p = p, m.sim = m.sim, prereturns, resids = t(sapply(res, FUN = function(x) x)), if(!is.null(mexsimdata)) t(sapply(mexsimdata, FUN = function(x) x)) else NULL)
for(i in 1:m.sim) seriesSim[[i]] = tmp[i, , drop = FALSE]
} else{
if( !is.null(cluster) ){
# when the GARCH model is without mean equation, simX == simulated residuals
clusterEvalQ(cluster, require(rmgarch))
clusterExport(cluster, c("Data", "varcoef", "p", "n.sim",
"n.start", "prereturns", "res", "mexsimdata"), envir = environment())
seriesSim = parLapply(cluster, as.list(1:m.sim), fun = function(i){
varxsim(X = Data, varcoef, p = p, n.sim = n.sim,
n.start = n.start, prereturns = prereturns,
resids = res[[i]], mexsimdata = mexsimdata[[i]])
})
} else{
seriesSim = lapply(as.list(1:m.sim), FUN = function(i){
varxsim(X = Data, varcoef, p = p, n.sim = n.sim,
n.start = n.start, prereturns = prereturns,
resids = res[[i]], mexsimdata = mexsimdata[[i]])
})
}
}
return( seriesSim )
}
mvmean.arsim = function(model, Data, res, arcoef, mxn, mexdata, mexsimdata,
prereturns, preresiduals, m.sim, n.sim, n.start, startMethod = "sample",
cluster = NULL)
{
seriesSim = vector(mode = "list", length = m.sim)
m = NCOL(Data)
if(model$modelinc[4]>0){
if( !is.null(mexsimdata) ){
if( !is.list(mexsimdata) | length(mexsimdata) != m.sim )
stop("\ngogarchsim-->error: mexsimdata must be a list of length equal to m.sim...", call. = FALSE)
for(j in 1:m.sim){
if( !is.matrix(mexsimdata[[j]]) )
stop("\ngogarchsim-->error: mexsimdata list submatrix must be a matrix...", call. = FALSE)
if( dim(mexsimdata[[j]])[2] != model$modelinc[4] )
stop("\ngogarchsim-->error: wrong number of external regressors in list submatrix!...", call. = FALSE)
if( dim(mexsimdata[[j]])[1] < (n.sim + n.start) )
stop("\ngogarchsim-->error: external regressor list submatrix has less points than requested simulation length (n.sim + n.start)...", call. = FALSE)
}
} else{
mexsimdata = vector(mode = "list", length = m.sim)
for(j in 1:m.sim) mexsimdata[[j]] = matrix(0, ncol = model$modelinc[4], nrow = (n.sim + n.start))
}
} else{
mexsimdata = NULL
}
p = model$modelinc[2]
if(!is.na(prereturns)[1]){
if(!is.matrix(prereturns))
stop("\nprereturns must be a matrix\n")
if(p>0) prereturns = tail(prereturns, p)
#if(dim(prereturns)[1] != p) stop("\nwrong number of rows for prereturns matrix")
if(p>0) if(NCOL(prereturns) != m)
stop("\nwrong number of cols for prereturns matrix")
} else{
if(startMethod == "sample"){
if(p>0) prereturns = as.matrix( tail(Data, p) )
} else{
if(p>0) prereturns = matrix(0, nrow = p, ncol = m, byrow = TRUE)
}
}
if(!is.na(preresiduals)[1]){
if(!is.matrix(preresiduals))
stop("\npreresiduals must be a matrix\n")
if(p>0) preresiduals = tail(preresiduals, p)
if(p>0) if(NCOL(preresiduals) != m)
stop("\nwrong number of cols for prereturns matrix")
} else{
if(startMethod != "sample")
if(p>0) preresiduals = matrix(0, nrow = p, ncol = m, byrow = TRUE)
}
if(model$modelinc[4]>0){
ex = mexdata[1:T, , drop = FALSE]
} else{
ex = NULL
}
if( !is.null(cluster) ){
tres = matrix(tail(preresiduals, p), ncol = m)
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("p", "ex", "arcoef", "mexsimdata",
"res", "n.start", "tres", "n.sim", "m.sim",
"prereturns"), envir = environment())
arxsim = parLapply(cluster, as.list(1:m), fun = function(i){
specx = rugarch::arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = rugarch::arfimapath(spec = specx, n.sim = n.sim,
n.start = n.start, m.sim = m.sim,
prereturns = if(p>0) tail(prereturns[,i],p) else NA,
preresiduals = if(p>0) tres[,i] else NA,
custom.dist = list(name = "sample",
distfit = matrix(sapply(res, FUN = function(x) x[,i]),
ncol = m.sim, nrow = n.sim, byrow = FALSE),
type = "res"),
mexsimdata = mexsimdata)
return(ans)
})
} else{
arxsim = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimapath(spec = specx, n.sim = n.sim, n.start = n.start,
m.sim = m.sim, prereturns = if(p>0) tail(prereturns[,i], p) else NA,
preresiduals = if(p>0) tail(preresiduals[,i], p) else NA,
custom.dist = list(name = "sample",
distfit = matrix(sapply(res, FUN = function(x) x[,i]),
ncol = m.sim, nrow = n.sim, byrow = FALSE),
type = "res"),
mexsimdata = mexsimdata)
return(ans)
})
}
# for(i in 1:m.sim) seriesSim[[i]] = matrix(sapply(arxsim, FUN = function(x) x@path$seriesSim[,i,drop=FALSE]), ncol = m, nrow = n.sim, byrow = TRUE)
for(i in 1:m.sim) seriesSim[[i]] = do.call(cbind, lapply(1:length(arxsim), function(j) arxsim[[j]]@path$seriesSim[,i,drop=FALSE]))
return( seriesSim )
}
Try the rmgarch package in your browser
Any scripts or data that you put into this service are public.
rmgarch documentation built on Feb. 5, 2022, 1:07 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions mvmean.arsim mvmean.varsim mvmean.arforecast mvmean.varforecast mvmean.arfilter mvmean.varfilter mvmean.arfit mvmean.varfit
#################################################################################
##
## R package rmgarch by Alexios Galanos Copyright (C) 2008-2022.
## This file is part of the R package rmgarch.
##
## The R package rmgarch is free software: you can redistribute it and/or modify
## it under the terms of the GNU General Public License as published by
## the Free Software Foundation, either version 3 of the License, or
## (at your option) any later version.
##
## The R package rmgarch is distributed in the hope that it will be useful,
## but WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
## GNU General Public License for more details.
##
#################################################################################
# First Stage Conditional Mean Filter
# ggn = goggarch numbering (+2 over copula and dcc models)
mvmean.varfit = function(model, data, VAR.fit = NULL, T, out.sample = 0, cluster = NULL, ggn = 0)
{
if( model$modelinc[2+ggn]>0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] ) stop("\ngogarchfit-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
if( !is.null(VAR.fit) ){
zdata = VAR.fit$xresiduals
# should be N - p - n.start
model$modelinc[1+ggn] = p = mp = VAR.fit$lag
N = dim(zdata)[1]
if(N != T) stop(paste("\ngogarchfit-->error: Residuals of VAR.fit object should be of length ", T, " (N - out.sample) ", sep=""))
mu = VAR.fit$xfitted
varcoef = VAR.fit$Bcoef
} else{
ic = model$varmodel$lag.criterion[1]
if( !is.null(model$varmodel$lag.max) ){
mp = .varxselect(y = data[1:T, ], lag.max = model$varmodel$lag.max, exogen = ex[1:T, , drop=FALSE])$selection
mp = as.numeric( mp[which(substr(names(mp), 1, 2) == substr(ic, 1, 2))] )
} else {
mp = model$modelinc[1+ggn]
}
p = mp
varmodel = varxfit(X = data[1:T, ,drop=FALSE], p = mp, constant = TRUE, exogen = ex[1:T, , drop=FALSE],
robust = model$varmodel$robust, gamma = model$varmodel$robust.control$gamma,
delta = model$varmodel$robust.control$delta, nc = model$varmodel$robust.control$nc,
ns = model$varmodel$robust.control$ns, postpad = "constant", cluster = cluster)
zdata = varmodel$xresiduals
# should be T
N = dim(zdata)[1]
mu = varmodel$xfitted
model$modelinc[1+ggn] = mp
varcoef = varmodel$Bcoef
}
# Filter Full Dataset now
if(out.sample > 0){
varmodel2 = varxfilter(X = data, p = mp, exogen = ex, Bcoef = varcoef, postpad = "constant")
zdata = varmodel2$xresiduals
# should be T - p + out.sample
N = dim(zdata)[1]
mu = varmodel2$xfitted
} else{
zdata = zdata
}
model$varcoef = varcoef
return(list(model = model, zdata = zdata, mu = mu, varcoef = varcoef, p = p, N = N))
}
# arfit not used by dcc and copula models
mvmean.arfit = function(model, data, ARcoef = NULL, T, out.sample = 0, solver,
solver.control = list(), fit.control = list(), cluster = NULL)
{
m = NCOL(data)
if(is.null(ARcoef)){
if( model$modelinc[4] > 0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfit-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
p = model$modelinc[2]
# Irrespsective of the Distribution, the mean filtration is always
# Normal.
arspec = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex[1:T, , drop = FALSE]),
distribution.model = "norm")
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("T", "arspec", "data", "solver",
"solver.control", "fit.control"), envir = environment())
arfit = parLapply(cluster, as.list(1:m), fun = function(i){
rugarch::arfimafit(spec = arspec, data = data[1:T,i,drop=FALSE],
out.sample = 0, solver = solver,
solver.control = solver.control,
fit.control = fit.control)
})
} else{
arfit = lapply(1:m, FUN = function(i) arfimafit(spec = arspec,
data = data[1:T,i,drop=FALSE], out.sample = 0,
solver = solver, solver.control = solver.control,
fit.control = fit.control))
}
zdata = sapply(arfit, FUN = function(x) residuals(x))
#cat("done!\n")
# rugarch always returns full length data
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
# should be T - p
N = NROW(zdata)
mu = sapply(arfit, FUN = function(x) fitted(x))
#if( p > 0 ) mu = mu[-c(1:p), , drop = FALSE]
arcoef = sapply(arfit, FUN = function(x) coef(x))
} else{
if( model$modelinc[4] > 0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfit-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
p = model$modelinc[2]
# Irrespsective of the Distribution, the mean filtration is always Normal.
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("T", "ARcoef", "data", "solver",
"solver.control", "fit.control"), envir = environment())
arfit = parLapply(cluster, as.list(1:m), fun = function(i){
arspec = rugarch::arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex[1:T, , drop = FALSE]),
distribution.model = "norm", fixed.pars = as.list(ARcoef[,i]));
ans = rugarch::arfimafilter(spec = arspec, data = data[1:T,i,drop=FALSE],
out.sample = 0);
return(ans)
})
} else{
arfit = lapply(1:m, FUN = function(i){
arspec = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex[1:T, , drop = FALSE]),
distribution.model = "norm", fixed.pars = as.list(ARcoef[,i]));
ans = arfimafilter(spec = arspec, data = data[1:T,i,drop=FALSE], out.sample = 0);
return(ans)
})
}
zdata = sapply(arfit, FUN = function(x) residuals(x))
#cat("done!\n")
# rugarch always returns full length data
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
# should be T - p
N = NROW(zdata)
mu = sapply(arfit, FUN = function(x) fitted(x))
#if( p > 0 ) mu = mu[-c(1:p), , drop = FALSE]
arcoef = ARcoef
}
if(out.sample>0){
if( !is.null(cluster) ){
clusterExport(cluster, c("data", "p", "ex", "arcoef"),
envir = environment())
arfit2 = parLapply(cluster, as.list(1:m), fun = function(i) {
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]));
ans = arfimafilter(spec = specx, data = data[,i, drop=FALSE], out.sample = 0);
return(ans)
})
} else{
arfit2 = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex), distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]));
ans = arfimafilter(spec = specx, data = data[,i, drop = FALSE], out.sample = 0);
return(ans)
})
}
zdata = sapply(arfit2, FUN = function(x) residuals(x))
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
N = NROW(zdata)
} else{
zdata = zdata
}
model$arcoef = arcoef
return(list(model = model, zdata = zdata, mu = mu, arcoef = arcoef, p = p, N = N))
}
mvmean.varfilter = function(model, data, varcoef, T, out.sample = 0, ggn = 0)
{
if( model$modelinc[2+ggn]>0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfilter-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
if(is.null(varcoef))
stop("\ncgarchfilter-->error: VAR model requires varcoef in specification to be supplied.")
p = model$modelinc[1+ggn]
vrmodel = varxfilter(X = data[1:T, , drop=FALSE], p = p, exogen = ex[1:T, , drop=FALSE], Bcoef = varcoef, postpad = "constant")
#cat("done!\n")
zdata = vrmodel$xresiduals
# should be N - p
N = NROW(data)
mu = vrmodel$xfitted
# if we have out of sample data we need to filter
if(out.sample > 0){
varmodel2 = varxfilter(X = data, p = p, exogen = ex, Bcoef = varcoef, postpad = "constant")
zdata = varmodel2$xresiduals
N = NROW(zdata)
mu = varmodel2$xfitted
} else{
zdata = zdata
}
model$varcoef = varcoef
model$residuals = zdata
return(list(model = model, zdata = zdata, mu = mu, p = p, N = N))
}
mvmean.arfilter = function(model, data, arcoef, T, out.sample = 0, cluster = NULL)
{
if( model$modelinc[4] > 0 ){
ex = model$modeldata$mexdata
if( dim(ex)[1] < dim(data)[1] )
stop("\ngogarchfilter-->error: external regressor matrix has less points than data!...", call. = FALSE)
} else{
ex = NULL
}
m = NCOL(data)
p = model$modelinc[2]
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("T", "data", "p", "ex", "arcoef"),
envir = environment())
armodel = parLapply(cluster, as.list(1:m), fun = function(i) {
specx = rugarch::arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex[1:T, ]),
distribution.model = "norm",
fixed.pars = as.list(arcoef[, i]))
ans = rugarch::arfimafilter(spec = specx, data = data[1:T, i], out.sample = 0)
return(ans)
})
} else{
armodel = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0), include.mean = TRUE,
arfima = FALSE, external.regressors = ex[1:T, ]), distribution.model = "norm",
fixed.pars = as.list(arcoef[, i]))
ans = arfimafilter(spec = specx, data = data[1:T, i], out.sample = 0)
return(ans)
})
}
#cat("done!\n")
zdata = sapply(armodel, FUN = function(x) residuals(x))
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
# should be N - p
N = NROW(data)
mu = sapply(armodel, FUN = function(x) fitted(x))
#if( p > 0 ) mu = mu[-c(1:p), , drop = FALSE]
# if we have out of sample data we need to filter
if(out.sample > 0){
if( !is.null(cluster) ){
clusterExport(cluster, c("data", "p", "ex", "arcoef"),
envir = environment())
armodel2 = parLapply(cluster, as.list(1:m), fun = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimafilter(spec = specx, data = data[,i], out.sample = 0)
return(ans)
})
} else{
armodel2 = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimafilter(spec = specx, data = data[,i], out.sample = 0)
return(ans)
})
}
zdata = sapply(armodel2, FUN = function(x) residuals(x))
#if( p > 0 ) zdata = zdata[-c(1:p), , drop = FALSE]
} else{
zdata = zdata
}
return(list(zdata = zdata, mu = mu, p = p, N = N))
}
mvmean.varforecast = function(model, mregfor, varcoef, n.ahead, n.roll, ns, ggn = 0){
m = NCOL(model$modeldata$data)
tf = n.ahead + n.roll
if( model$modelinc[2+ggn] > 0 ){
if( is.null( mregfor ) ){
warning("\nExternal Regressor Forecasts Matrix NULL...setting to zero...\n")
mregfor = matrix(0, ncol = model$modelinc[2+ggn], nrow = (n.roll + n.ahead) )
} else{
if( dim(mregfor)[2] != model$modelinc[2+ggn] )
stop("\ngogarchforecast-->error: wrong number of external regressors!...", call. = FALSE)
if( dim(mregfor)[1] < (n.roll + n.ahead) )
stop("\ngogarchforecast-->error: external regressor matrix has less points than requested forecast length (1+n.roll) x n.ahead!...", call. = FALSE)
}
} else{
mregfor = NULL
}
if( n.ahead == 1 && (n.roll > ns) )
stop("\ngogarchforecast-->error: n.roll greater than out.sample!", call. = FALSE)
# varxforecast returns array --> change to matrix since we do not allow
# mixed n.roll and n.ahead
Mu = varxforecast(X = model$modeldata$data, Bcoef = varcoef, p = model$modelinc[1+ggn], out.sample = ns, n.ahead, n.roll, mregfor)
return(Mu)
}
mvmean.arforecast = function(model, mregfor, arcoef, n.ahead, n.roll, ns, cluster = NULL)
{
m = NCOL(model$modeldata$data)
tf = n.ahead + n.roll
if( !is.null(cluster) ){
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("model", "arcoef", "mregfor", "ns",
"n.ahead", "n.roll"), envir = environment())
arfx = parLapply(cluster, as.list(1:m), fun = function(i){
specx = rugarch::arfimaspec(mean.model = list(armaOrder = c(model$modelinc[2], 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = model$modeldata$mexdata),
distribution.model = "norm", fixed.pars = as.list(arcoef[,i]))
ans = rugarch::arfimaforecast(fitORspec = specx, data = model$modeldata$data[,i],
out.sample = ns, n.ahead = n.ahead, n.roll = n.roll,
external.forecasts = list(mregfor = mregfor))
return(ans)
})
} else{
arfx = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(model$modelinc[2], 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = model$modeldata$mexdata),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimaforecast(fitORspec = specx, data = model$modeldata$data[,i],
out.sample = ns, n.ahead = n.ahead, n.roll = n.roll,
external.forecasts = list(mregfor = mregfor))
return(ans)
})
}
Mu = array(NA, dim=c(n.ahead, m, n.roll+1))
for(i in 1:(n.roll+1)){
Mu[,,i] = matrix(sapply(arfx, FUN = function(x) fitted(x)[,i], simplify=TRUE), ncol = m)
}
return(Mu)
}
mvmean.varsim = function(model, Data, res, mexsimdata, prereturns, m.sim,
n.sim, n.start, startMethod = "sample", cluster = NULL, ggn = 0)
{
seriesSim = vector(mode = "list", length = m.sim)
m = NCOL(Data)
varcoef = model$varcoef
if( model$modelinc[2+ggn] > 0 ){
if( !is.null(mexsimdata) ){
if( !is.list(mexsimdata) | length(mexsimdata) != m.sim )
stop("\ngogarchsim-->error: mexsimdata must be a list of length equal to m.sim...", call. = FALSE)
for(j in 1:m.sim){
if( !is.matrix(mexsimdata[[j]]) )
stop("\ngogarchsim-->error: mexsimdata list submatrix must be a matrix...", call. = FALSE)
if( dim(mexsimdata[[j]])[2] != model$modelinc[2+ggn] )
stop("\ngogarchsim-->error: wrong number of external regressors in list submatrix!...", call. = FALSE)
if( dim(mexsimdata[[j]])[1] < (n.sim + n.start) )
stop("\ngogarchsim-->error: external regressor list submatrix has less points than requested simulation length (n.sim + n.start)...", call. = FALSE)
}
} else{
mexsimdata = vector(mode = "list", length = m.sim)
for(j in 1:m.sim) mexsimdata[[j]] = matrix(0, ncol = model$modelinc[2+ggn], nrow = (n.sim + n.start))
}
} else{
mexsimdata = NULL
}
p = model$modelinc[1+ggn]
if(startMethod == "sample"){
if( !is.null(prereturns) ){
if(!is.matrix(prereturns))
stop("\nprereturns must be a matrix\n")
if(dim(prereturns)[1] < p)
stop("\nwrong number of rows for prereturns matrix")
if(dim(prereturns)[2] != m)
stop("\nwrong number of cols for prereturns matrix")
prereturns = matrix(tail(prereturns, p), nrow = p, ncol = m)
} else{
prereturns = as.matrix( tail(Data, p) )
}
} else{
prereturns = matrix(0, nrow = p, ncol = m, byrow = TRUE)
}
if( n.sim == 1 && n.start == 0 ){
# SPECIAL ROUTINE TO SPEED UP 1-AHEAD ESTIMATION
tmp = varxsimXX(X = Data, varcoef, p = p, m.sim = m.sim, prereturns, resids = t(sapply(res, FUN = function(x) x)), if(!is.null(mexsimdata)) t(sapply(mexsimdata, FUN = function(x) x)) else NULL)
for(i in 1:m.sim) seriesSim[[i]] = tmp[i, , drop = FALSE]
} else{
if( !is.null(cluster) ){
# when the GARCH model is without mean equation, simX == simulated residuals
clusterEvalQ(cluster, require(rmgarch))
clusterExport(cluster, c("Data", "varcoef", "p", "n.sim",
"n.start", "prereturns", "res", "mexsimdata"), envir = environment())
seriesSim = parLapply(cluster, as.list(1:m.sim), fun = function(i){
varxsim(X = Data, varcoef, p = p, n.sim = n.sim,
n.start = n.start, prereturns = prereturns,
resids = res[[i]], mexsimdata = mexsimdata[[i]])
})
} else{
seriesSim = lapply(as.list(1:m.sim), FUN = function(i){
varxsim(X = Data, varcoef, p = p, n.sim = n.sim,
n.start = n.start, prereturns = prereturns,
resids = res[[i]], mexsimdata = mexsimdata[[i]])
})
}
}
return( seriesSim )
}
mvmean.arsim = function(model, Data, res, arcoef, mxn, mexdata, mexsimdata,
prereturns, preresiduals, m.sim, n.sim, n.start, startMethod = "sample",
cluster = NULL)
{
seriesSim = vector(mode = "list", length = m.sim)
m = NCOL(Data)
if(model$modelinc[4]>0){
if( !is.null(mexsimdata) ){
if( !is.list(mexsimdata) | length(mexsimdata) != m.sim )
stop("\ngogarchsim-->error: mexsimdata must be a list of length equal to m.sim...", call. = FALSE)
for(j in 1:m.sim){
if( !is.matrix(mexsimdata[[j]]) )
stop("\ngogarchsim-->error: mexsimdata list submatrix must be a matrix...", call. = FALSE)
if( dim(mexsimdata[[j]])[2] != model$modelinc[4] )
stop("\ngogarchsim-->error: wrong number of external regressors in list submatrix!...", call. = FALSE)
if( dim(mexsimdata[[j]])[1] < (n.sim + n.start) )
stop("\ngogarchsim-->error: external regressor list submatrix has less points than requested simulation length (n.sim + n.start)...", call. = FALSE)
}
} else{
mexsimdata = vector(mode = "list", length = m.sim)
for(j in 1:m.sim) mexsimdata[[j]] = matrix(0, ncol = model$modelinc[4], nrow = (n.sim + n.start))
}
} else{
mexsimdata = NULL
}
p = model$modelinc[2]
if(!is.na(prereturns)[1]){
if(!is.matrix(prereturns))
stop("\nprereturns must be a matrix\n")
if(p>0) prereturns = tail(prereturns, p)
#if(dim(prereturns)[1] != p) stop("\nwrong number of rows for prereturns matrix")
if(p>0) if(NCOL(prereturns) != m)
stop("\nwrong number of cols for prereturns matrix")
} else{
if(startMethod == "sample"){
if(p>0) prereturns = as.matrix( tail(Data, p) )
} else{
if(p>0) prereturns = matrix(0, nrow = p, ncol = m, byrow = TRUE)
}
}
if(!is.na(preresiduals)[1]){
if(!is.matrix(preresiduals))
stop("\npreresiduals must be a matrix\n")
if(p>0) preresiduals = tail(preresiduals, p)
if(p>0) if(NCOL(preresiduals) != m)
stop("\nwrong number of cols for prereturns matrix")
} else{
if(startMethod != "sample")
if(p>0) preresiduals = matrix(0, nrow = p, ncol = m, byrow = TRUE)
}
if(model$modelinc[4]>0){
ex = mexdata[1:T, , drop = FALSE]
} else{
ex = NULL
}
if( !is.null(cluster) ){
tres = matrix(tail(preresiduals, p), ncol = m)
clusterEvalQ(cluster, loadNamespace("rugarch"))
clusterExport(cluster, c("p", "ex", "arcoef", "mexsimdata",
"res", "n.start", "tres", "n.sim", "m.sim",
"prereturns"), envir = environment())
arxsim = parLapply(cluster, as.list(1:m), fun = function(i){
specx = rugarch::arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = rugarch::arfimapath(spec = specx, n.sim = n.sim,
n.start = n.start, m.sim = m.sim,
prereturns = if(p>0) tail(prereturns[,i],p) else NA,
preresiduals = if(p>0) tres[,i] else NA,
custom.dist = list(name = "sample",
distfit = matrix(sapply(res, FUN = function(x) x[,i]),
ncol = m.sim, nrow = n.sim, byrow = FALSE),
type = "res"),
mexsimdata = mexsimdata)
return(ans)
})
} else{
arxsim = lapply(1:m, FUN = function(i){
specx = arfimaspec(mean.model = list(armaOrder = c(p, 0),
include.mean = TRUE, arfima = FALSE,
external.regressors = ex),
distribution.model = "norm",
fixed.pars = as.list(arcoef[,i]))
ans = arfimapath(spec = specx, n.sim = n.sim, n.start = n.start,
m.sim = m.sim, prereturns = if(p>0) tail(prereturns[,i], p) else NA,
preresiduals = if(p>0) tail(preresiduals[,i], p) else NA,
custom.dist = list(name = "sample",
distfit = matrix(sapply(res, FUN = function(x) x[,i]),
ncol = m.sim, nrow = n.sim, byrow = FALSE),
type = "res"),
mexsimdata = mexsimdata)
return(ans)
})
}
# for(i in 1:m.sim) seriesSim[[i]] = matrix(sapply(arxsim, FUN = function(x) x@path$seriesSim[,i,drop=FALSE]), ncol = m, nrow = n.sim, byrow = TRUE)
for(i in 1:m.sim) seriesSim[[i]] = do.call(cbind, lapply(1:length(arxsim), function(j) arxsim[[j]]@path$seriesSim[,i,drop=FALSE]))
return( seriesSim )
}
Try the rmgarch package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.