Nothing
R/spec_rjd.R
In RJDemetra: Interface to 'JDemetra+' Seasonal Adjustment Software
Defines functions
specSeats_r2jd
specX11_r2jd
spec_TRAMO_r2jd
spec_regarima_X13_r2jd
arimaCoef_jd2r
arimaCoef_r2jd
preVar_r2jd
preOut_r2jd
span_r2jd
specSeats_jd2r
specX11_jd2r
spec_TRAMO_jd2r
spec_regarima_X13_jd2r
jd_span
# Functions to extract information from the java object
jd_span <- function(type,d0,d1,n0,n1){
x <- if (type=="All") {"All"} else if (type=="From") {paste("From",d0, sep=" ")}
else if (type=="To") {paste("Until",d1, sep=" ")}
else if (type=="Between") {paste(d0,d1,sep=" - ")}
else if (type=="First") {paste("All but first",n0,"periods", sep=" ")}
else if (type=="Last") {paste("All but last",n1,"periods", sep=" ")}
else if (type=="Excluding") {paste("All but first",n0,"periods and last",n1,"periods", sep=" ")}
return(x)
}
spec_regarima_X13_jd2r <- function(spec = NA, context_dictionary = NULL,
extra_info = FALSE,
freq = NA){
#Estimate
preliminary.check <- spec$getBasic()$isPreliminaryCheck()
jestimate <-.jcall(spec,"Ljdr/spec/x13/EstimateSpec;","getEstimate")
jest.span <-.jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
estimate.type <- .jcall(jest.span,"S","getType")
estimate.d0 <- .jcall(jest.span,"S","getD0")
estimate.d1 <- .jcall(jest.span,"S","getD1")
estimate.n0 <- .jcall(jest.span,"I","getN0")
estimate.n1 <- .jcall(jest.span,"I","getN1")
estimate.span <- jd_span(type= estimate.type,d0=estimate.d0,d1=estimate.d1,n0=estimate.n0,n1=estimate.n1)
if (estimate.type %in% c("All", "First", "Last", "Excluding")) {
estimate.d0 <- estimate.d1 <- NA
} else if (estimate.type=="From") {
estimate.d0 <- NA
} else if (estimate.type=="To") {
estimate.d1 <- NA
}
estimate.tol <-.jcall(jestimate ,"D","getTol")
#Transform
jtransform <-.jcall(spec,"Ljdr/spec/x13/TransformSpec;","getTransform")
transform.function<-.jcall(jtransform,"S","getFunction")
transform.adjust<-.jcall(jtransform,"S","getAdjust")
transform.aicdiff<-.jcall(jtransform,"D","getAic")
#Regression
jregression<-.jcall(spec,"Ljdr/spec/x13/RegressionSpec;","getRegression")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/x13/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/x13/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/x13/EasterSpec;","getEaster")
tradingdays.option <- .jcall(jtd,"S","getOption")
if(tradingdays.option != "UserDefined"){
tradingdays.option <- .jcall(jtd,"S","getTradingDays")
}
tradingdays.autoadjust <- .jcall(jtd,"Z","isAutoAdjust")
tradingdays.leapyear <- .jcall(jtd,"S","getLengthOfPeriod")
tradingdays.stocktd <- .jcall(jtd,"I","getW")
tradingdays.test <- .jcall(jtd,"S","getTest")
easter.enabled <- .jcall(jeaster,"Z","isEnabled")
easter.julian <- .jcall(jeaster,"Z","isJulian")
easter.duration <- .jcall(jeaster,"I","getDuration")
easter.test <- .jcall(jeaster,"S","getTest")
#Outlier
joutlier <- .jcall(spec,"Ljdr/spec/x13/OutlierSpec;","getOutliers")
joutlier.span <- .jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
outlier.enabled <- .jcall(joutlier,"Z","isEnabled")
outlier.type <- .jcall(joutlier.span,"S","getType")
outlier.d0 <- .jcall(joutlier.span,"S","getD0")
outlier.d1 <- .jcall(joutlier.span,"S","getD1")
outlier.n0 <- .jcall(joutlier.span,"I","getN0")
outlier.n1 <- .jcall(joutlier.span,"I","getN1")
outlier.span <- jd_span(type = outlier.type,
d0 = outlier.d0, d1 = outlier.d1,
n0 = outlier.n0, n1 = outlier.n1)
if (outlier.type %in% c("All", "First", "Last", "Excluding")) {
outlier.d0 <- outlier.d1 <- NA
} else if (outlier.type=="From") {
outlier.d0 <- NA
} else if (outlier.type=="To") {
outlier.d1 <- NA
}
outlier.ao <- .jcall(joutlier, "Z", "isAO")
outlier.tc <- .jcall(joutlier, "Z", "isTC")
outlier.ls <- .jcall(joutlier, "Z", "isLS")
outlier.so <- .jcall(joutlier, "Z", "isSO")
outlier.usedefcv <- .jcall(joutlier, "Z", "isDefaultVa")
outlier.cv <- .jcall(joutlier, "D", "getVa")
outlier.method <- .jcall(joutlier, "S", "getMethod")
outlier.tcrate <- .jcall(joutlier, "D", "getTCRate")
#Arima
jarima <- .jcall(spec, "Ljdr/spec/x13/ArimaSpec;", "getArima")
automdl.enabled <- .jcall(jarima, "Z", "isAmiEnabled")
automdl.acceptdefault <- .jcall(jarima,"Z","isAcceptDefault")
automdl.cancel <- .jcall(jarima,"D","getCancel")
automdl.ub1 <- .jcall(jarima,"D","getUb1")
automdl.ub2 <- .jcall(jarima,"D","getUb2")
automdl.mixed <- .jcall(jarima,"Z","isMixed")
automdl.balanced <- .jcall(jarima,"Z","isBalanced")
automdl.armalimit <- .jcall(jarima,"D","getTsig")
automdl.reducecv <- .jcall(jarima,"D","getPredCV")
automdl.ljungboxlimit <- .jcall(jarima,"D","getPcr")
automdl.ubfinal <- .jcall(jarima,"D","getUbFinal")
arima.mu <- .jcall(jarima,"Z","isMean")
arima.p <- .jcall(jarima,"I","getP")
arima.d <- .jcall(jarima,"I","getD")
arima.q <- .jcall(jarima,"I","getQ")
arima.bp <- .jcall(jarima,"I","getBP")
arima.bd <- .jcall(jarima,"I","getBD")
arima.bq <- .jcall(jarima,"I","getBQ")
#span matrix
type <- c(estimate.type,outlier.type)
d0 <- c(estimate.d0,outlier.d0)
d1 <- c(estimate.d1,outlier.d1)
n0 <- c(estimate.n0,outlier.n0)
n1 <- c(estimate.n1,outlier.n1)
span <- data.frame(type=type,d0=d0,d1=d1,n0=n0,n1=n1)
rownames(span) <- c("estimate","outlier")
#Default values:
arima.coef <- FALSE
userdef_spec <- list(specification = data.frame(outlier = FALSE,
outlier.coef = FALSE,
variables = FALSE,
variables.coef = FALSE,
stringsAsFactors = FALSE),
outliers = NA, variables = list(series = NA, description = NA))
arima.coef.spec <- NA
result <- list(preliminary.check = preliminary.check, estimate.type = estimate.type, estimate.d0 = estimate.d0, estimate.d1 = estimate.d1,
estimate.n0 = estimate.n0 , estimate.n1 = estimate.n1, estimate.span = estimate.span, estimate.tol = estimate.tol,
transform.function = transform.function, transform.adjust = transform.adjust, transform.aicdiff = transform.aicdiff,
tradingdays.option = tradingdays.option , tradingdays.autoadjust = tradingdays.autoadjust,
tradingdays.leapyear = tradingdays.leapyear , tradingdays.stocktd = tradingdays.stocktd, tradingdays.test = tradingdays.test,
easter.enabled = easter.enabled , easter.julian = easter.julian , easter.duration = easter.duration, easter.test = easter.test,
outlier.enabled = outlier.enabled, outlier.type = outlier.type, outlier.d0 = outlier.d0, outlier.d1 = outlier.d1,
outlier.n0 = outlier.n0, outlier.n1 = outlier.n1, outlier.span = outlier.span, outlier.ao = outlier.ao,
outlier.tc = outlier.tc, outlier.ls = outlier.ls, outlier.so = outlier.so, outlier.usedefcv = outlier.usedefcv,
outlier.cv = outlier.cv, outlier.method = outlier.method, outlier.tcrate = outlier.tcrate,
automdl.enabled = automdl.enabled, automdl.acceptdefault = automdl.acceptdefault, automdl.cancel = automdl.cancel,
automdl.ub1 = automdl.ub1, automdl.ub2 = automdl.ub2, automdl.mixed = automdl.mixed , automdl.balanced = automdl.balanced ,
automdl.armalimit = automdl.armalimit, automdl.reducecv = automdl.reducecv, automdl.ljungboxlimit = automdl.ljungboxlimit,
automdl.ubfinal = automdl.ubfinal, arima.mu = arima.mu, arima.p = arima.p, arima.d = arima.d, arima.q = arima.q,
arima.bp = arima.bp, arima.bd = arima.bd, arima.bq = arima.bq, span = span,
arima.coef = arima.coef,arima.coef.spec = arima.coef.spec,
userdef_spec = userdef_spec)
# Extra info importing from a workspace
if(!extra_info)
return(result)
n_prespecified_out <- .jcall(jregression,"I","getPrespecifiedOutliersCount")
if(n_prespecified_out > 0 ){
result$userdef_spec$specification$outlier <- TRUE
outliers <- lapply(1:n_prespecified_out, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$Outlier;",
"getPrespecifiedOutlier",
as.integer(i-1))
})
type <- sapply(outliers, function(x) x$getCode())
date <- sapply(outliers, function(x) x$getPosition())
coeff <- sapply(outliers, function(x) x$getCoefficient())
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$outlier.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$outlier.coef <- TRUE
}
outlier_spec <- data.frame(type = type, date = date, coeff = coeff)
result$userdef_spec$outliers <- outlier_spec
}
n_userdefined_var <- .jcall(jregression,"I","getUserDefinedVariablesCount")
if(n_userdefined_var > 0 ){
result$userdef_spec$specification$variables <- TRUE
ud_vars <- lapply(1:n_userdefined_var, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$UserDefinedVariable;",
"getUserDefinedVariable",
as.integer(i-1))
})
type <- sapply(ud_vars, function(x) x$getComponent())
coeff <- sapply(ud_vars, function(x) x$getCoefficient())
var_names <- sapply(ud_vars, function(x) x$getName())
var_names_split <- strsplit(var_names,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
var_names <- base::make.names(var_names, unique = TRUE)
var_names <- gsub(".","_", var_names, fixed = TRUE)
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$variables.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$variables.coef <- TRUE
}
result$userdef_spec$variables$description <- data.frame(type = type,
coeff = coeff,
row.names = var_names)
# To check variables group names
# t <- context_dictionary$getTsVariableManagers()
# t$getNames()
#
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
#Calendar
user_td <- jtd$getUserVariables()
if(length(user_td) > 0 ){
var_names_split <- strsplit(user_td,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
var_names <- base::make.names(var_names, unique = TRUE)
var_names <- gsub(".","_", var_names, fixed = TRUE)
result$userdef_spec$specification$variables <-
TRUE
td_var_description <- data.frame(type = rep("Calendar",length(var_names)),
coeff = NA, row.names = var_names)
if(identical_na(result$userdef_spec$variables$description)){
result$userdef_spec$variables$description <- td_var_description
}else{
result$userdef_spec$variables$description <- rbind(result$userdef_spec$variables$description,
td_var_description)
}
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(!identical_na(result$userdef_spec$variables$series)){
var_series <- ts.union(result$userdef_spec$variables$series, var_series)
}
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
## Ramp effects
core_regression <- jregression$getCore()$getRegression()
jramps <- core_regression$getRamps()
nb_ramps <- core_regression$getRampsCount()
if (nb_ramps > 0) {
var_names <- sapply(seq_len(nb_ramps), function(x) jramps[[x]]$getDescription())
result$userdef_spec$specification$variables <-
TRUE
coeff <- NA
if (core_regression$hasFixedCoefficients()) {
coeff <- sapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
ramp_name <- jramp$getName()
fixed_coeff <- core_regression$getFixedCoefficients(ramp_name)
if (is.null(fixed_coeff)) {
NA
}else{
fixed_coeff
}
})
result$userdef_spec$specification$variables.coef <- TRUE
}
td_var_description <- data.frame(type = rep("Series", length(var_names)),
coeff = coeff, row.names = var_names)
if (identical_na(result$userdef_spec$variables$description)) {
result$userdef_spec$variables$description <-
td_var_description
}else{
result$userdef_spec$variables$description <-
rbind(result$userdef_spec$variables$description,
td_var_description)
}
if (!is.na(freq) || !identical_na(result$userdef_spec$variables$series)) {
if (is.na(freq))
freq <- frequency(result$userdef_spec$variables$series)
if (!identical_na(result$userdef_spec$variables$series)) {
start_ts <- start(result$userdef_spec$variables$series)
end_ts <- end(result$userdef_spec$variables$series)
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start = start_ts, end = end_ts,
start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
} else {
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
}
ramp_series <- ts(simplify2array(ramp_series),
start = start(ramp_series[[1]]),
frequency = frequency(ramp_series[[1]]))
if (!identical_na(result$userdef_spec$variables$series)) {
ramp_series <- ts.union(result$userdef_spec$variables$series, ramp_series)
}
if (is.mts(ramp_series))
colnames(ramp_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- ramp_series
}
}
Phi <- jarima$getPhi()
BPhi <- jarima$getBPhi()
Theta <- jarima$getTheta()
BTheta <- jarima$getBTheta()
arima_coefficients_spec <-
rbind(arimaCoef_jd2r(Phi),
arimaCoef_jd2r(BPhi),
arimaCoef_jd2r(Theta),
arimaCoef_jd2r(BTheta))
if(!is.null(arima_coefficients_spec) &&
any(arima_coefficients_spec$Type != "Undefined")){
result$arima.coef <- TRUE
result$arima.coef.spec <- arima_coefficients_spec
}
result
}
spec_TRAMO_jd2r <- function(spec = NA, context_dictionary = NULL,
extra_info = FALSE, freq = NA){
#Estimate
preliminary.check <- spec$getBasic()$isPreliminaryCheck()
jestimate <- .jcall(spec,"Ljdr/spec/tramoseats/EstimateSpec;","getEstimate")
jest.span <- .jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
estimate.type <- .jcall(jest.span,"S","getType")
estimate.d0 <- .jcall(jest.span,"S","getD0")
estimate.d1 <- .jcall(jest.span,"S","getD1")
estimate.n0 <- .jcall(jest.span,"I","getN0")
estimate.n1 <- .jcall(jest.span,"I","getN1")
estimate.span <- jd_span(type= estimate.type,d0=estimate.d0,d1=estimate.d1,n0=estimate.n0,n1=estimate.n1)
if (estimate.type %in% c("All", "First", "Last", "Excluding")) {
estimate.d0 <- estimate.d1 <- NA
} else if (estimate.type=="From") {
estimate.d0 <- NA
} else if (estimate.type=="To") {
estimate.d1 <- NA
}
estimate.tol <-.jcall(jestimate ,"D","getTol")
estimate.eml <-.jcall(jestimate ,"Z","isEml")
estimate.urfinal <-.jcall(jestimate ,"D","getUbp")
#Transform
jtransform <-.jcall(spec,"Ljdr/spec/tramoseats/TransformSpec;","getTransform")
transform.function <-.jcall(jtransform,"S","getFunction")
transform.fct <-.jcall(jtransform,"D","getFct")
#Regression
jregression<-.jcall(spec,"Ljdr/spec/tramoseats/RegressionSpec;","getRegression")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/tramoseats/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/tramoseats/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/tramoseats/EasterSpec;","getEaster")
tradingdays.mauto <-.jcall(jtd,"S","getAutomatic")
tradingdays.pftd <-.jcall(jtd,"D","getPftd")
tradingdays.option <- .jcall(jtd,"S","getOption")
if(tradingdays.option != "UserDefined"){
tradingdays.option <- .jcall(jtd,"S","getTradingDays")
}
tradingdays.leapyear <-.jcall(jtd,"Z","getLeapYear")
tradingdays.stocktd <-.jcall(jtd,"I","getW")
tradingdays.test <-.jcall(jtd,"S","getRegressionTestType")
easter.type <- .jcall(jeaster,"S","getOption")
easter.julian <-.jcall(jeaster,"Z","isJulian")
easter.duration <-.jcall(jeaster,"I","getDuration")
easter.test <-.jcall(jeaster,"Z","isTest")
#Outlier
joutlier<-.jcall(spec,"Ljdr/spec/tramoseats/OutlierSpec;","getOutlier")
joutlier.span <-.jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
outlier.enabled <-.jcall(joutlier,"Z","isOutliersDetectionEnabled")
outlier.type <- .jcall(joutlier.span,"S","getType")
outlier.d0 <- .jcall(joutlier.span,"S","getD0")
outlier.d1 <- .jcall(joutlier.span,"S","getD1")
outlier.n0 <- .jcall(joutlier.span,"I","getN0")
outlier.n1 <- .jcall(joutlier.span,"I","getN1")
outlier.span <- jd_span(type= outlier.type,d0=outlier.d0,d1=outlier.d1,n0=outlier.n0,n1=outlier.n1)
if (outlier.type %in% c("All", "First", "Last", "Excluding")) {
outlier.d0 <- outlier.d1 <- NA
} else if (outlier.type=="From") {
outlier.d0 <- NA
} else if (outlier.type=="To") {
outlier.d1 <- NA
}
outlier.ao <-.jcall(joutlier,"Z","isAO")
outlier.tc <-.jcall(joutlier,"Z","isTC")
outlier.ls <-.jcall(joutlier,"Z","isLS")
outlier.so <-.jcall(joutlier,"Z","isSO")
outlier.usedefcv <-.jcall(joutlier,"Z","isAutoVa")
outlier.cv <-.jcall(joutlier,"D","getVa")
outlier.eml <-.jcall(joutlier,"Z","isEML")
outlier.tcrate <-.jcall(joutlier,"D","getTCRate")
#Arima
jarima<-.jcall(spec,"Ljdr/spec/tramoseats/ArimaSpec;","getArima")
automdl.enabled <-.jcall(jarima,"Z","isEnabled")
automdl.acceptdefault <-.jcall(jarima,"Z","isAcceptDefault")
automdl.cancel <-.jcall(jarima,"D","getCancel")
automdl.ub1 <-.jcall(jarima,"D","getUb1")
automdl.ub2 <-.jcall(jarima,"D","getUb2")
automdl.armalimit <-.jcall(jarima,"D","getTsig")
automdl.reducecv <-.jcall(jarima,"D","getPc")
automdl.ljungboxlimit <-.jcall(jarima,"D","getPcr")
automdl.compare <-.jcall(jarima,"Z","isAmiCompare")
arima.mu <-.jcall(jarima,"Z","isMean")
arima.p <-.jcall(jarima,"I","getP")
arima.d <-.jcall(jarima,"I","getD")
arima.q <-.jcall(jarima,"I","getQ")
arima.bp <-.jcall(jarima,"I","getBP")
arima.bd <-.jcall(jarima,"I","getBD")
arima.bq <-.jcall(jarima,"I","getBQ")
#span matrix
type<-c(estimate.type,outlier.type)
d0<-c(estimate.d0,outlier.d0)
d1<-c(estimate.d1,outlier.d1)
n0<-c(estimate.n0,outlier.n0)
n1<-c(estimate.n1,outlier.n1)
span<-data.frame(type=type,d0=d0,d1=d1,n0=n0,n1=n1)
rownames(span)<-c("estimate","outlier")
arima.coef <- FALSE
userdef_spec <-list(specification = data.frame(outlier = FALSE,
outlier.coef = FALSE,
variables = FALSE,
variables.coef = FALSE,
stringsAsFactors = FALSE),
outliers = NA, variables = list(series = NA, description = NA))
arima.coef.spec <- NA
result <- list(preliminary.check = preliminary.check,
estimate.type = estimate.type,estimate.d0 = estimate.d0,estimate.d1 = estimate.d1,estimate.n0 = estimate.n0,
estimate.n1 = estimate.n1,estimate.span = estimate.span,estimate.tol = estimate.tol,estimate.eml = estimate.eml,
estimate.urfinal = estimate.urfinal,transform.function = transform.function,transform.fct = transform.fct,
tradingdays.mauto = tradingdays.mauto,tradingdays.pftd = tradingdays.pftd,tradingdays.option = tradingdays.option,
tradingdays.leapyear = tradingdays.leapyear,tradingdays.stocktd = tradingdays.stocktd,
tradingdays.test = tradingdays.test,easter.type = easter.type,easter.julian = easter.julian,easter.duration = easter.duration,
easter.test = easter.test,outlier.enabled = outlier.enabled,outlier.type = outlier.type,outlier.d0 = outlier.d0,
outlier.d1 = outlier.d1,outlier.n0 = outlier.n0,outlier.n1 = outlier.n1,outlier.span = outlier.span,outlier.ao = outlier.ao,
outlier.tc = outlier.tc,outlier.ls = outlier.ls,outlier.so = outlier.so,outlier.usedefcv = outlier.usedefcv,
outlier.cv = outlier.cv,outlier.eml = outlier.eml,outlier.tcrate = outlier.tcrate,automdl.enabled = automdl.enabled,
automdl.acceptdefault = automdl.acceptdefault,automdl.cancel = automdl.cancel,automdl.ub1 = automdl.ub1,
automdl.ub2 = automdl.ub2,automdl.armalimit = automdl.armalimit,automdl.reducecv = automdl.reducecv,
automdl.ljungboxlimit = automdl.ljungboxlimit,automdl.compare = automdl.compare,arima.mu = arima.mu,arima.p = arima.p,
arima.d = arima.d,arima.q = arima.q,arima.bp = arima.bp,arima.bd = arima.bd,arima.bq = arima.bq, span = span,
arima.coef = arima.coef,arima.coef.spec = arima.coef.spec,
userdef_spec = userdef_spec)
# Extra info importing from a workspace
if(!extra_info)
return(result)
n_prespecified_out <- .jcall(jregression,"I","getPrespecifiedOutliersCount")
if(n_prespecified_out > 0 ){
# Outlier.coef is set to TRUE: if the coefficient isn't fixed, the
# coef value will be equal to 0 and the outlier will be estimated
result$userdef_spec$specification$outlier <- TRUE
outliers <- lapply(1:n_prespecified_out, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$Outlier;",
"getPrespecifiedOutlier",
as.integer(i-1))
})
type <- sapply(outliers, function(x) x$getCode())
date <- sapply(outliers, function(x) x$getPosition())
coeff <- sapply(outliers, function(x) x$getCoefficient())
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$outlier.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$outlier.coef <- TRUE
}
outlier_spec <- data.frame(type = type, date = date, coeff = coeff)
result$userdef_spec$outliers <- outlier_spec
}
n_userdefined_var <- .jcall(jregression,"I","getUserDefinedVariablesCount")
if(n_userdefined_var > 0 ){
# variables.coef is set to TRUE: if the coefficient isn't fixed, the
# coef value will be equal to 0 and the variable will be estimated
result$userdef_spec$specification$variables <- TRUE
ud_vars <- lapply(1:n_userdefined_var, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$UserDefinedVariable;",
"getUserDefinedVariable",
as.integer(i-1))
})
type <- sapply(ud_vars, function(x) x$getComponent())
coeff <- sapply(ud_vars, function(x) x$getCoefficient())
var_names <- sapply(ud_vars, function(x) x$getName())
var_names_split <- strsplit(var_names,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$variables.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$variables.coef <- TRUE
}
result$userdef_spec$variables$description <- data.frame(type = type,
coeff = coeff,
row.names = var_names)
# To check variables group names
# t <- context_dictionary$getTsVariableManagers()
# t$getNames()
#
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
#Calendar
user_td <- jtd$getUserVariables()
if(length(user_td) > 0 ){
var_names_split <- strsplit(user_td,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
result$userdef_spec$specification$variables <-
TRUE
coeff <- NA
# if (core_regression$hasFixedCoefficients()) {
# # coeff <- sapply(seq_len(nb_ramps), function(i){
# # jramp <- jramps[[i]]
# # ramp_name <- jramp$getName()
# # fixed_coeff <- core_regression$getFixedCoefficients(ramp_name)
# # if (is.null(fixed_coeff)) {
# # NA
# # }else{
# # fixed_coeff
# # }
# # })
# # result$userdef_spec$specification$variables.coef <- TRUE
# }
td_var_description <- data.frame(type = rep("Calendar",length(var_names)),
coeff = coeff, row.names = var_names)
if(identical_na(result$userdef_spec$variables$description)){
result$userdef_spec$variables$description <- td_var_description
}else{
result$userdef_spec$variables$description <- rbind(result$userdef_spec$variables$description,
td_var_description)
}
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(!identical_na(result$userdef_spec$variables$series)){
var_series <- ts.union(result$userdef_spec$variables$series, var_series)
}
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
## Ramp effects
core_regression <- jregression$getCore()$getRegression()
jramps <- core_regression$getRamps()
nb_ramps <- core_regression$getRampsCount()
if (nb_ramps > 0) {
var_names <- sapply(seq_len(nb_ramps), function(x) jramps[[x]]$getDescription())
result$userdef_spec$specification$variables <-
TRUE
coeff <- NA
if (core_regression$hasFixedCoefficients()) {
coeff <- sapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
ramp_name <- jramp$getName()
fixed_coeff <- core_regression$getFixedCoefficients(ramp_name)
if (is.null(fixed_coeff)) {
NA
}else{
fixed_coeff
}
})
result$userdef_spec$specification$variables.coef <- TRUE
}
td_var_description <- data.frame(type = rep("Series", length(var_names)),
coeff = coeff, row.names = var_names)
if (identical_na(result$userdef_spec$variables$description)) {
result$userdef_spec$variables$description <-
td_var_description
}else{
result$userdef_spec$variables$description <-
rbind(result$userdef_spec$variables$description,
td_var_description)
}
if (!is.na(freq) || !identical_na(result$userdef_spec$variables$series)) {
if (is.na(freq))
freq <- frequency(result$userdef_spec$variables$series)
if (!identical_na(result$userdef_spec$variables$series)) {
start_ts <- start(result$userdef_spec$variables$series)
end_ts <- end(result$userdef_spec$variables$series)
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start = start_ts, end = end_ts,
start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
} else {
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
}
ramp_series <- ts(simplify2array(ramp_series),
start = start(ramp_series[[1]]),
frequency = frequency(ramp_series[[1]]))
if (!identical_na(result$userdef_spec$variables$series)) {
ramp_series <- ts.union(result$userdef_spec$variables$series, ramp_series)
}
if (is.mts(ramp_series))
colnames(ramp_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- ramp_series
}
}
#Arima
Phi <- jarima$getPhi()
BPhi <- jarima$getBPhi()
Theta <- jarima$getTheta()
BTheta <- jarima$getBTheta()
arima_coefficients_spec <-
rbind(arimaCoef_jd2r(Phi),
arimaCoef_jd2r(BPhi),
arimaCoef_jd2r(Theta),
arimaCoef_jd2r(BTheta))
if(!is.null(arima_coefficients_spec) &&
any(arima_coefficients_spec$Type != "Undefined")){
result$arima.coef <- TRUE
result$arima.coef.spec <- arima_coefficients_spec
}
result
}
specX11_jd2r <- function(spec = NA, freq = NA){
jx11 <- .jcall(spec,"Ljdr/spec/x13/X11Spec;","getX11")
if (!is.na(freq)) {
.jcall(jx11,"V","setFreq", as.integer(freq))
fullseasonalma <- .jcall(jx11,"[S","getFullSeasonalMA")
if (!is.null(fullseasonalma) && length(unique(fullseasonalma)) > 1) {
seasonalma <- paste(fullseasonalma, collapse = ", ")
}else{
seasonalma <- .jcall(jx11,"S","getSeasonalMA")
}
}else{
seasonalma <- .jcall(jx11,"S","getSeasonalMA")
}
mode <- .jcall(jx11,"S","getMode")
seasonalComp <- .jcall(jx11,"Z","isSeasonal")
lsigma <- .jcall(jx11,"D","getLSigma")
usigma <- .jcall(jx11,"D","getUSigma")
trendAuto <- .jcall(jx11,"Z","isAutoTrendMA")
trendma <- .jcall(jx11,"I","getTrendMA")
fcasts <- .jcall(jx11,"I","getForecastHorizon")
bcasts <- .jcall(jx11,"I","getBackcastHorizon")
excludeFcasts <- .jcall(jx11,"Z","isExcludefcst")
calendarSigma <- jx11$getCalendarSigma()$toString()
sigmaVector <- .jcall(jx11,returnSig = "[S", "getSigmavec") #TODO
if (length(sigmaVector) == 0) {
sigmaVector <- NA
}else{
sigmaVector <- paste(sigmaVector, collapse = ", ")
}
var <- data.frame(mode,seasonalComp,lsigma,usigma,trendAuto,trendma,seasonalma,
fcasts, bcasts,
calendarSigma, sigmaVector,
excludeFcasts,
stringsAsFactors = FALSE)
names(var) <- sprintf("x11.%s",
c("mode","seasonalComp","lsigma","usigma",
"trendAuto","trendma","seasonalma",
"fcasts","bcasts",
"calendarSigma", "sigmaVector",
"excludeFcasts"))
return(var)
}
specSeats_jd2r <- function(spec = NA){
jseats <- .jcall(spec,"Ljdr/spec/tramoseats/SeatsSpec;","getSeats")
predictionLength <- .jcall(jseats,"I","getPredictionLength")
approx <- .jcall(jseats,"S","getApproximationMode")
maBoundary <- .jcall(jseats,"D","getXl")
trendBoundary <- .jcall(jseats,"D","getRMod")
seasdBoundary <- .jcall(jseats,"D","getSMod")
seasdBoundary1 <- .jcall(jseats,"D","getSMod1")
seasTol <- .jcall(jseats,"D","getEpsPhi")
method <- .jcall(jseats,"S","getMethod")
var <- list(predictionLength, approx,maBoundary,trendBoundary,seasdBoundary,seasdBoundary1,seasTol,
method)
names(var) <- c("predictionLength", "approx", "maBoundary", "trendBoundary",
"seasdBoundary","seasdBoundary1","seasTol",
"method")
return(var)
}
# Functions to introduce modifications in the java object
span_r2jd <- function(jsobjct = NA, type = NA, d0=NA, d1=NA, n0 = NA, n1=NA){
if (type =="All") {
.jcall(jsobjct,"V","all")
} else if (type=="From"){
.jcall(jsobjct,"V","from",d0)
} else if (type=="To"){
.jcall(jsobjct,"V","to",d1)
} else if (type=="Between"){
.jcall(jsobjct,"V","between",d0,d1)
} else if (type=="First"){
.jcall(jsobjct,"V","first",n0)
} else if (type=="Last"){
.jcall(jsobjct,"V","last",n1)
} else if (type=="Excluding"){
.jcall(jsobjct,"V","excluding",n0,n1)
}
}
preOut_r2jd <- function(jsobjct = NA, coefEna = NA, out = NA, outDate = NA, outCoef = NA){
coef <- if (coefEna == TRUE) {outCoef} else {rep(0,length(out))}
for (i in 1:length(out)){
.jcall(jsobjct,"V","addPrespecifiedOutlier", out[i],outDate[i],coef[i])
}
}
preVar_r2jd <- function(jsobjct = NA, jsdict = NA, coefEna = NA,
prevar_spec = list(series = NA, description = data.frame(NA,NA)),
jtd = NA) {
series = prevar_spec$series
varType = prevar_spec$description[,1]
varCoef = prevar_spec$description[,2]
varNames <- rownames(prevar_spec$description)
nvar <- if (is.mts(series)) {dim(series)[2]} else if (is.ts(series)) {1} else {0}
type <- if (all(is.na(varType))) {rep("Undefined",nvar)} else {as.character(varType)}
coef <- if (coefEna == TRUE) {varCoef} else {rep(0,nvar)}
if(nvar == 0)
return(NULL)
calendar_def <- grep("Calendar",type)
n_calendar_def <- length(calendar_def)
jcoreg = jsobjct$getCore()$getRegression()
if (nvar == 1){
if(n_calendar_def == 1){
.jcall(jsdict,"V","add",varNames,ts_r2jd(series))
if(coef!=0){
jcoreg$setFixedCoefficients(paste0("td|r@",varNames),
.jarray(coef))
}
.jcall(jtd,"V","setUserVariables", .jarray(paste0("r.",varNames)))
}else{
.jcall(jsdict,"V","add",varNames,ts_r2jd(series))
.jcall(jsobjct,"V","addUserDefinedVariable",varNames,type[1],coef[1])
}
}else{
if(n_calendar_def == 0 | n_calendar_def == nvar){
if(n_calendar_def >0){
for (i in 1:nvar){
.jcall(jsdict,"V","add",varNames[i],ts_r2jd(series[,i]))
if(coef[i]!=0){
jcoreg$setFixedCoefficients(paste0("td|r@",varNames[i]),
.jarray(coef[i]))
}
}
.jcall(jtd,"V","setUserVariables",
paste0("r.",varNames)
)
}else{
for (i in 1:nvar){
.jcall(jsdict,"V","add",varNames[i],ts_r2jd(series[,i]))
.jcall(jsobjct,"V","addUserDefinedVariable",varNames[i],type[i],coef[i])
}
}
}else{
i_ud <- (1:nvar)[-calendar_def]
for (i in i_ud){
.jcall(jsdict,"V","add", varNames[i],
ts_r2jd(series[, i]))
.jcall(jsobjct,"V","addUserDefinedVariable", varNames[i],
type[i], coef[i])
}
for (i in calendar_def){
.jcall(jsdict,"V","add",varNames[i],
ts_r2jd(series[, i]))
if(coef[i]!=0){
jcoreg$setFixedCoefficients(paste0("td|r@",varNames[i]),
.jarray(coef[i]))
}
}
.jcall(jtd,"V","setUserVariables",
.jarray(paste0("r.",varNames[calendar_def])))
}
}
}
arimaCoef_r2jd <- function(jsobjct = NA, acoef = NA, p = NA , q = NA, bp = NA, bq = NA){
typ <- acoef[,1]
val <- acoef[,2]
np <- length(typ)
par <- val
par[is.na(val) | val==0 | typ=="Undefined"] <- NaN
fix<-rep(FALSE,np)
fix[typ=="Fixed" & !is.na(par)] <- TRUE
n <-1
if (p!=0){
param <- parameters_r2jd(par[n:(n+p-1)],fix[n:(n+p-1)])
.jcall(jsobjct,"V","setPhi",param)
n <- n+p
}
if (q!=0){
param <- parameters_r2jd(par[n:(n+q-1)],fix[n:(n+q-1)])
.jcall(jsobjct,"V","setTheta",param)
n <- n+q
}
if (bp!=0){
param <- parameters_r2jd(par[n:(n+bp-1)],fix[n:(n+bp-1)])
.jcall(jsobjct,"V","setBPhi",param)
n <- n+bp
}
if (bq!=0){
param <- parameters_r2jd(par[n:(n+bq-1)],fix[n:(n+bq-1)])
.jcall(jsobjct,"V","setBTheta",param)
}
}
arimaCoef_jd2r <- function(jparams){
if (is.jnull(jparams))
return(NULL)
param<-.jcastToArray(jparams)
len <- length(param)
if (len==0)
return (NULL)
param_name <- deparse(substitute(jparams))
Type <- sapply(param, function(x) x$getType()$toString())
Value <- sapply(param, function(x) x$getValue())
data_param <- data.frame(Type = Type, Value = Value)
rownames(data_param) <- sprintf("%s(%i)",
param_name,
1:len)
data_param
}
spec_regarima_X13_r2jd <- function(spec = NA, jdspec = NA){
if (is.null(s_estimate(spec)))
return(.jnew("jdr/spec/ts/Utility$Dictionary"))
est <- s_estimate(spec)
trans <- s_transform(spec)
usrspc <- s_usrdef(spec)
outF <- s_preOut(spec)
varF <- s_preVar(spec)
td <- s_td(spec)
easter <- s_easter(spec)
outliers <- s_out(spec)
arimaspc <- s_arima(spec)
arimacoF <- s_arimaCoef(spec)
span <- s_span(spec)
#Estimate
jestimate <-.jcall(jdspec,"Ljdr/spec/x13/EstimateSpec;","getEstimate")
jest.span <-.jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = jest.span, type = span[1,1], d0=as.character(span[1,2]), d1=as.character(span[1,3]),
n0 = as.integer(span[1,4]), n1=as.integer(span[1,5]))
.jcall(jestimate ,"V","setTol", as.numeric(est["tolerance"]))
jdspec$getBasic()$setPreliminaryCheck(est[1, "preliminary.check"])
#Transform
jtransform <-.jcall(jdspec,"Ljdr/spec/x13/TransformSpec;","getTransform")
.jcall(jtransform,"V","setFunction",as.character(trans[1]))
.jcall(jtransform,"V","setAdjust",as.character(trans[2]))
.jcall(jtransform,"V","setAic",as.numeric(trans[3]))
#Regression
jregression<-.jcall(jdspec,"Ljdr/spec/x13/RegressionSpec;","getRegression")
#Pre-specified outliers
.jcall(jregression,"V","clearPrespecifiedOutliers")
if (usrspc[1]==TRUE)
preOut_r2jd(jsobjct = jregression, coefEna= usrspc[2], out= as.character(outF[,1]),outDate= as.character(outF[,2]),outCoef= outF[,3])
#User-defined variables
.jcall(jregression,"V","clearUserDefinedVariables")
jdictionary <- .jnew("jdr/spec/ts/Utility$Dictionary")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/x13/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/x13/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/x13/EasterSpec;","getEaster")
.jcall(jeaster,"V","setTest", as.character(easter[4]))
.jcall(jeaster,"V","setJulian",as.logical(easter[2]))
.jcall(jeaster,"V","setDuration", as.integer(easter[3]))
.jcall(jeaster,"V","setEnabled",as.logical(easter[1]))
#Calendar options:
if(td[1] == "UserDefined"){
.jcall(jtd,"V","setOption","UserDefined")
}else{
if (td[4]==0) {
.jcall(jtd,"V","setW",as.integer(td[4]))
.jcall(jtd,"V","setTradingDays",as.character(td[1]))
}else{
.jcall(jtd,"V","setTradingDays",as.character(td[1]))
.jcall(jtd,"V","setW",as.integer(td[4]))
}
.jcall(jtd,"V","setAutoAdjust",as.logical(td[2]))
.jcall(jtd,"V","setLengthOfPeriod", as.character(td[3]))
}
.jcall(jtd,"V","setTest",as.character(td[5]))
#user-defined specification
if (usrspc[3]==TRUE)
preVar_r2jd(jsobjct = jregression, jsdict = jdictionary, coefEna = usrspc[4],
prevar_spec = varF, jtd = jtd)
#Outlier
joutlier<-.jcall(jdspec,"Ljdr/spec/x13/OutlierSpec;","getOutliers")
joutlier.span <-.jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = joutlier.span, type = span[2,1], d0=as.character(span[2,2]), d1=as.character(span[2,3]),
n0 = as.integer(span[2,4]), n1=as.integer(span[2,5]))
.jcall(joutlier,"V","setAO",as.logical(outliers[3]))
.jcall(joutlier,"V","setTC",as.logical(outliers[4]))
.jcall(joutlier,"V","setLS",as.logical(outliers[5]))
.jcall(joutlier,"V","setSO",as.logical(outliers[6]))
# Default critical value
if (outliers[7] == TRUE){
.jcall(joutlier,"V","setDefaultVa",as.logical(outliers[7]))
}else{
.jcall(joutlier,"V","setVa",as.numeric(outliers[8]))
}
.jcall(joutlier,"V","setMethod",as.character(outliers[9]))
.jcall(joutlier,"V","setTCRate",as.numeric(outliers[10]))
.jcall(joutlier,"V","setEnabled",as.logical(outliers[1]))
#Arima
jarima<-.jcall(jdspec,"Ljdr/spec/x13/ArimaSpec;","getArima")
.jcall(jarima,"V","setAmiEnabled",as.logical(arimaspc[1]))
.jcall(jarima,"V","setAcceptDefault",as.logical(arimaspc[2]))
.jcall(jarima,"V","setCancel",as.numeric(arimaspc[3]))
.jcall(jarima,"V","setUb1", as.numeric(arimaspc[4]))
.jcall(jarima,"V","setUb2", as.numeric(arimaspc[5]))
.jcall(jarima,"V","setMixed", as.logical(arimaspc[6]))
.jcall(jarima,"V","setBalanced", as.logical(arimaspc[7]))
.jcall(jarima,"V","setTsig",as.numeric(arimaspc[8]))
.jcall(jarima,"V","setPredCV",as.numeric(arimaspc[9]))
.jcall(jarima,"V","setPcr", as.numeric(arimaspc[10]))
.jcall(jarima,"V","setUbFinal", as.numeric(arimaspc[11]))
.jcall(jarima,"V","setMean",as.logical(arimaspc[12]))
.jcall(jarima,"V","setP",as.integer(arimaspc[13]))
.jcall(jarima,"V","setD",as.integer(arimaspc[14]))
.jcall(jarima,"V","setQ",as.integer(arimaspc[15]))
.jcall(jarima,"V","setBP",as.integer(arimaspc[16]))
.jcall(jarima,"V","setBD",as.integer(arimaspc[17]))
.jcall(jarima,"V","setBQ",as.integer(arimaspc[18]))
# Fixed ARIMA coefficients
if (arimaspc[19]==TRUE)
arimaCoef_r2jd(jsobjct = jarima, acoef = arimacoF, p = as.numeric(arimaspc[13]) , q = as.numeric(arimaspc[15]),
bp = as.numeric(arimaspc[16]), bq = as.numeric(arimaspc[18]))
return(jdictionary)
}
spec_TRAMO_r2jd <- function(spec = NA, jdspec =NA){
est <- s_estimate(spec)
trans <- s_transform(spec)
usrspc <- s_usrdef(spec)
outF <- s_preOut(spec)
varF <- s_preVar(spec)
td <- s_td(spec)
easter <- s_easter(spec)
outliers <- s_out(spec)
arimaspc <- s_arima(spec)
arimacoF <- s_arimaCoef(spec)
span <- s_span(spec)
#Estimate
jdspec$getBasic()$setPreliminaryCheck(est[1, "preliminary.check"])
jestimate <-.jcall(jdspec,"Ljdr/spec/tramoseats/EstimateSpec;","getEstimate")
jest.span <-.jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = jest.span, type = span[1,1], d0=as.character(span[1,2]), d1=as.character(span[1,3]),
n0 = as.integer(span[1,4]), n1=as.integer(span[1,5]))
.jcall(jestimate ,"V","setTol",as.numeric(est["tolerance"]))
.jcall(jestimate ,"V","setEml",as.logical(est["exact_ml"]))
.jcall(jestimate ,"V","setUbp",as.numeric(est["urfinal"]))
#Transform
jtransform <-.jcall(jdspec,"Ljdr/spec/tramoseats/TransformSpec;","getTransform")
.jcall(jtransform,"V","setFunction",as.character(trans[1]))
.jcall(jtransform,"V","setFct", as.numeric(trans[2]))
#Regression
jregression<-.jcall(jdspec,"Ljdr/spec/tramoseats/RegressionSpec;","getRegression")
#Pre-specified outliers
.jcall(jregression,"V","clearPrespecifiedOutliers")
if (usrspc[1]==TRUE)
preOut_r2jd(jsobjct = jregression, coefEna= usrspc[2],out= as.character(outF[,1]), outDate= as.character(outF[,2]),outCoef= outF[,3])
#User-defined variables
.jcall(jregression,"V","clearUserDefinedVariables")
jdictionary <- .jnew("jdr/spec/ts/Utility$Dictionary")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/tramoseats/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/tramoseats/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/tramoseats/EasterSpec;","getEaster")
.jcall(jtd,"V","setAutomatic", as.character(td[1]))
.jcall(jtd,"V","setPftd", as.numeric(td[2]))
if(td[3] == "UserDefined"){
.jcall(jtd,"V","setOption","UserDefined")
}else{
if (td[5]==0) {
# .jcall(jtd,"V","setW",as.integer(td[5]))
.jcall(jtd,"V","setTradingDays",as.character(td[3]))
}else{
.jcall(jtd,"V","setTradingDays",as.character(td[3]))
.jcall(jtd,"V","setW",as.integer(td[5]))
}
.jcall(jtd,"V","setLeapYear",as.logical(td[4]))
}
.jcall(jtd,"V","setRegressionTestType", as.character(td[6]))
.jcall(jeaster,"V","setOption",as.character(easter[1]))
.jcall(jeaster,"V","setJulian",as.logical(easter[2]))
.jcall(jeaster,"V","setDuration",as.integer(easter[3]))
.jcall(jeaster,"V","setTest",as.logical(easter[4]))
#user-defined specification
if (usrspc[3]==TRUE)
preVar_r2jd(jsobjct = jregression, jsdict = jdictionary, coefEna = usrspc[4],
prevar_spec = varF, jtd = jtd)
#Outlier
joutlier<-.jcall(jdspec,"Ljdr/spec/tramoseats/OutlierSpec;","getOutlier")
joutlier.span <-.jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = joutlier.span, type = span[2,1], d0=as.character(span[2,2]), d1=as.character(span[2,3]),
n0 = as.integer(span[2,4]), n1=as.integer(span[2,5]))
.jcall(joutlier,"V","setOutliersDetectionEnabled", as.logical(outliers[1]))
.jcall(joutlier,"V","setAO", as.logical(outliers[3]))
.jcall(joutlier,"V","setTC", as.logical(outliers[4]))
.jcall(joutlier,"V","setLS", as.logical(outliers[5]))
.jcall(joutlier,"V","setSO", as.logical(outliers[6]))
# Default critical value
if (outliers[7] == TRUE){
.jcall(joutlier,"V","setAutoVa", as.logical(outliers[7]))
}else{
.jcall(joutlier,"V","setVa", as.numeric(outliers[8]))
}
.jcall(joutlier,"V","setEML", as.logical(outliers[9]))
.jcall(joutlier,"V","setTCRate", as.numeric(outliers[10]))
#Arima
jarima<-.jcall(jdspec,"Ljdr/spec/tramoseats/ArimaSpec;","getArima")
.jcall(jarima,"V","setEnabled", as.logical(arimaspc[1]))
.jcall(jarima,"V","setAcceptDefault", as.logical(arimaspc[2]))
.jcall(jarima,"V","setCancel", as.numeric(arimaspc[3]))
.jcall(jarima,"V","setUb1", as.numeric(arimaspc[4]))
.jcall(jarima,"V","setUb2", as.numeric(arimaspc[5]))
.jcall(jarima,"V","setTsig", as.numeric(arimaspc[6]))
.jcall(jarima,"V","setPc", as.numeric(arimaspc[7]))
.jcall(jarima,"V","setPcr", as.numeric(arimaspc[8]))
.jcall(jarima,"V","setAmiCompare", as.logical(arimaspc[9]))
.jcall(jarima,"V","setMean", as.logical(arimaspc[10]))
.jcall(jarima,"V","setP", as.integer(arimaspc[11]))
.jcall(jarima,"V","setD", as.integer(arimaspc[12]))
.jcall(jarima,"V","setQ", as.integer(arimaspc[13]))
.jcall(jarima,"V","setBP", as.integer(arimaspc[14]))
.jcall(jarima,"V","setBD", as.integer(arimaspc[15]))
.jcall(jarima,"V","setBQ", as.integer(arimaspc[16]))
# Fixed ARIMA coefficients
if (arimaspc[17]==TRUE)
arimaCoef_r2jd(jsobjct = jarima, acoef = arimacoF, p = as.numeric(arimaspc[11]) , q = as.numeric(arimaspc[13]),
bp = as.numeric(arimaspc[14]), bq = as.numeric(arimaspc[16]))
return(jdictionary)
}
specX11_r2jd <- function(rspec = NA, jdspec = NA , freq = NA){
x11 <- s_x11(rspec)
jx11 <- .jcall(jdspec,"Ljdr/spec/x13/X11Spec;","getX11")
seasonalma <- unlist(strsplit(as.character(x11[["x11.seasonalma"]]),
split = ", "))
len.ma <- length(seasonalma)
.jcall(jx11,"V","setMode",as.character(x11[["x11.mode"]]))
.jcall(jx11,"V","setSeasonal",as.logical(x11[["x11.seasonalComp"]]))
.jcall(jx11,"V","setLSigma",as.numeric(x11[["x11.lsigma"]]))
.jcall(jx11,"V","setUSigma",as.numeric(x11[["x11.usigma"]]))
if (x11[["x11.trendAuto"]]) {
.jcall(jx11,"V","setAutoTrendMA",as.logical(TRUE))
}else{
.jcall(jx11,"V","setAutoTrendMA",as.logical(FALSE))
.jcall(jx11,"V","setTrendMA",as.integer(x11[["x11.trendma"]]))
}
.jcall(jx11,"V","setFreq", as.integer(freq))
if (len.ma == 1) {
.jcall(jx11, "V", "setSeasonalMA", seasonalma)
seasma <- seasonalma
}else if (len.ma != freq) {
.jcall(jx11,"V","setSeasonalMA","Msr")
warning(paste0("wrong frequency of the x11.seasonalma (",
len.ma, " instead of ", freq, ").",
"\nPre-specified seasonal filters will be ignored (x11.seasonalma=\"Msr\")."),
call. = FALSE)
seasma <- "Msr"
} else {
.jcall(jx11,"V","setFullSeasonalMA",seasonalma)
seasma <- as.character(x11[["x11.seasonalma"]])
}
.jcall(jx11, "V", "setForecastHorizon", as.integer(x11[["x11.fcasts"]]))
.jcall(jx11, "V", "setBackcastHorizon", as.integer(x11[["x11.bcasts"]]))
.jcall(jx11, "V", "setCalendarSigma", x11[["x11.calendarSigma"]])
if (x11[["x11.calendarSigma"]] == "Select" &&
!identical_na(x11[["x11.sigmaVector"]])) {
# sigmaVector is change only if x11.calendarSigma is set to "Select"
sigmaVector <- unlist(strsplit(as.character(x11[["x11.sigmaVector"]]),
split = ", "))
if (length(sigmaVector) != freq) {
warning(paste0("Wrong frequency of the x11.sigmaVector (",
length(sigmaVector), " instead of ", freq, ").",
"\nThis parameter will be ignored."), call. = FALSE)
} else {
.jcall(jx11, "V", "setSigmavec", sigmaVector)
}
}
.jcall(jx11, "V", "setExcludefcst", as.logical(x11[["x11.excludeFcasts"]]))
return(seasma)
}
specSeats_r2jd <- function(rspec = NA, jdspec = NA){
seats <- s_seats(rspec)
jseats <-.jcall(jdspec,"Ljdr/spec/tramoseats/SeatsSpec;","getSeats")
.jcall(jseats,"V","setPredictionLength", as.integer(seats[["seats.predictionLength"]]))
.jcall(jseats,"V","setApproximationMode", as.character(seats[["seats.approx"]]))
.jcall(jseats,"V","setXl", as.numeric(seats[["seats.maBoundary"]]))
.jcall(jseats,"V","setRMod", as.numeric(seats[["seats.trendBoundary"]]))
.jcall(jseats,"V","setSMod", as.numeric(seats[["seats.seasdBoundary"]]))
.jcall(jseats,"V","setSMod1", as.numeric(seats[["seats.seasdBoundary1"]]))
.jcall(jseats,"V","setEpsPhi",as.numeric(seats[["seats.seasTol"]]))
.jcall(jseats,"V","setMethod", as.character(seats[["seats.method"]]))
}
Try the RJDemetra package in your browser
Any scripts or data that you put into this service are public.
RJDemetra documentation built on Oct. 19, 2023, 1:12 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 specSeats_r2jd specX11_r2jd spec_TRAMO_r2jd spec_regarima_X13_r2jd arimaCoef_jd2r arimaCoef_r2jd preVar_r2jd preOut_r2jd span_r2jd specSeats_jd2r specX11_jd2r spec_TRAMO_jd2r spec_regarima_X13_jd2r jd_span
# Functions to extract information from the java object
jd_span <- function(type,d0,d1,n0,n1){
x <- if (type=="All") {"All"} else if (type=="From") {paste("From",d0, sep=" ")}
else if (type=="To") {paste("Until",d1, sep=" ")}
else if (type=="Between") {paste(d0,d1,sep=" - ")}
else if (type=="First") {paste("All but first",n0,"periods", sep=" ")}
else if (type=="Last") {paste("All but last",n1,"periods", sep=" ")}
else if (type=="Excluding") {paste("All but first",n0,"periods and last",n1,"periods", sep=" ")}
return(x)
}
spec_regarima_X13_jd2r <- function(spec = NA, context_dictionary = NULL,
extra_info = FALSE,
freq = NA){
#Estimate
preliminary.check <- spec$getBasic()$isPreliminaryCheck()
jestimate <-.jcall(spec,"Ljdr/spec/x13/EstimateSpec;","getEstimate")
jest.span <-.jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
estimate.type <- .jcall(jest.span,"S","getType")
estimate.d0 <- .jcall(jest.span,"S","getD0")
estimate.d1 <- .jcall(jest.span,"S","getD1")
estimate.n0 <- .jcall(jest.span,"I","getN0")
estimate.n1 <- .jcall(jest.span,"I","getN1")
estimate.span <- jd_span(type= estimate.type,d0=estimate.d0,d1=estimate.d1,n0=estimate.n0,n1=estimate.n1)
if (estimate.type %in% c("All", "First", "Last", "Excluding")) {
estimate.d0 <- estimate.d1 <- NA
} else if (estimate.type=="From") {
estimate.d0 <- NA
} else if (estimate.type=="To") {
estimate.d1 <- NA
}
estimate.tol <-.jcall(jestimate ,"D","getTol")
#Transform
jtransform <-.jcall(spec,"Ljdr/spec/x13/TransformSpec;","getTransform")
transform.function<-.jcall(jtransform,"S","getFunction")
transform.adjust<-.jcall(jtransform,"S","getAdjust")
transform.aicdiff<-.jcall(jtransform,"D","getAic")
#Regression
jregression<-.jcall(spec,"Ljdr/spec/x13/RegressionSpec;","getRegression")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/x13/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/x13/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/x13/EasterSpec;","getEaster")
tradingdays.option <- .jcall(jtd,"S","getOption")
if(tradingdays.option != "UserDefined"){
tradingdays.option <- .jcall(jtd,"S","getTradingDays")
}
tradingdays.autoadjust <- .jcall(jtd,"Z","isAutoAdjust")
tradingdays.leapyear <- .jcall(jtd,"S","getLengthOfPeriod")
tradingdays.stocktd <- .jcall(jtd,"I","getW")
tradingdays.test <- .jcall(jtd,"S","getTest")
easter.enabled <- .jcall(jeaster,"Z","isEnabled")
easter.julian <- .jcall(jeaster,"Z","isJulian")
easter.duration <- .jcall(jeaster,"I","getDuration")
easter.test <- .jcall(jeaster,"S","getTest")
#Outlier
joutlier <- .jcall(spec,"Ljdr/spec/x13/OutlierSpec;","getOutliers")
joutlier.span <- .jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
outlier.enabled <- .jcall(joutlier,"Z","isEnabled")
outlier.type <- .jcall(joutlier.span,"S","getType")
outlier.d0 <- .jcall(joutlier.span,"S","getD0")
outlier.d1 <- .jcall(joutlier.span,"S","getD1")
outlier.n0 <- .jcall(joutlier.span,"I","getN0")
outlier.n1 <- .jcall(joutlier.span,"I","getN1")
outlier.span <- jd_span(type = outlier.type,
d0 = outlier.d0, d1 = outlier.d1,
n0 = outlier.n0, n1 = outlier.n1)
if (outlier.type %in% c("All", "First", "Last", "Excluding")) {
outlier.d0 <- outlier.d1 <- NA
} else if (outlier.type=="From") {
outlier.d0 <- NA
} else if (outlier.type=="To") {
outlier.d1 <- NA
}
outlier.ao <- .jcall(joutlier, "Z", "isAO")
outlier.tc <- .jcall(joutlier, "Z", "isTC")
outlier.ls <- .jcall(joutlier, "Z", "isLS")
outlier.so <- .jcall(joutlier, "Z", "isSO")
outlier.usedefcv <- .jcall(joutlier, "Z", "isDefaultVa")
outlier.cv <- .jcall(joutlier, "D", "getVa")
outlier.method <- .jcall(joutlier, "S", "getMethod")
outlier.tcrate <- .jcall(joutlier, "D", "getTCRate")
#Arima
jarima <- .jcall(spec, "Ljdr/spec/x13/ArimaSpec;", "getArima")
automdl.enabled <- .jcall(jarima, "Z", "isAmiEnabled")
automdl.acceptdefault <- .jcall(jarima,"Z","isAcceptDefault")
automdl.cancel <- .jcall(jarima,"D","getCancel")
automdl.ub1 <- .jcall(jarima,"D","getUb1")
automdl.ub2 <- .jcall(jarima,"D","getUb2")
automdl.mixed <- .jcall(jarima,"Z","isMixed")
automdl.balanced <- .jcall(jarima,"Z","isBalanced")
automdl.armalimit <- .jcall(jarima,"D","getTsig")
automdl.reducecv <- .jcall(jarima,"D","getPredCV")
automdl.ljungboxlimit <- .jcall(jarima,"D","getPcr")
automdl.ubfinal <- .jcall(jarima,"D","getUbFinal")
arima.mu <- .jcall(jarima,"Z","isMean")
arima.p <- .jcall(jarima,"I","getP")
arima.d <- .jcall(jarima,"I","getD")
arima.q <- .jcall(jarima,"I","getQ")
arima.bp <- .jcall(jarima,"I","getBP")
arima.bd <- .jcall(jarima,"I","getBD")
arima.bq <- .jcall(jarima,"I","getBQ")
#span matrix
type <- c(estimate.type,outlier.type)
d0 <- c(estimate.d0,outlier.d0)
d1 <- c(estimate.d1,outlier.d1)
n0 <- c(estimate.n0,outlier.n0)
n1 <- c(estimate.n1,outlier.n1)
span <- data.frame(type=type,d0=d0,d1=d1,n0=n0,n1=n1)
rownames(span) <- c("estimate","outlier")
#Default values:
arima.coef <- FALSE
userdef_spec <- list(specification = data.frame(outlier = FALSE,
outlier.coef = FALSE,
variables = FALSE,
variables.coef = FALSE,
stringsAsFactors = FALSE),
outliers = NA, variables = list(series = NA, description = NA))
arima.coef.spec <- NA
result <- list(preliminary.check = preliminary.check, estimate.type = estimate.type, estimate.d0 = estimate.d0, estimate.d1 = estimate.d1,
estimate.n0 = estimate.n0 , estimate.n1 = estimate.n1, estimate.span = estimate.span, estimate.tol = estimate.tol,
transform.function = transform.function, transform.adjust = transform.adjust, transform.aicdiff = transform.aicdiff,
tradingdays.option = tradingdays.option , tradingdays.autoadjust = tradingdays.autoadjust,
tradingdays.leapyear = tradingdays.leapyear , tradingdays.stocktd = tradingdays.stocktd, tradingdays.test = tradingdays.test,
easter.enabled = easter.enabled , easter.julian = easter.julian , easter.duration = easter.duration, easter.test = easter.test,
outlier.enabled = outlier.enabled, outlier.type = outlier.type, outlier.d0 = outlier.d0, outlier.d1 = outlier.d1,
outlier.n0 = outlier.n0, outlier.n1 = outlier.n1, outlier.span = outlier.span, outlier.ao = outlier.ao,
outlier.tc = outlier.tc, outlier.ls = outlier.ls, outlier.so = outlier.so, outlier.usedefcv = outlier.usedefcv,
outlier.cv = outlier.cv, outlier.method = outlier.method, outlier.tcrate = outlier.tcrate,
automdl.enabled = automdl.enabled, automdl.acceptdefault = automdl.acceptdefault, automdl.cancel = automdl.cancel,
automdl.ub1 = automdl.ub1, automdl.ub2 = automdl.ub2, automdl.mixed = automdl.mixed , automdl.balanced = automdl.balanced ,
automdl.armalimit = automdl.armalimit, automdl.reducecv = automdl.reducecv, automdl.ljungboxlimit = automdl.ljungboxlimit,
automdl.ubfinal = automdl.ubfinal, arima.mu = arima.mu, arima.p = arima.p, arima.d = arima.d, arima.q = arima.q,
arima.bp = arima.bp, arima.bd = arima.bd, arima.bq = arima.bq, span = span,
arima.coef = arima.coef,arima.coef.spec = arima.coef.spec,
userdef_spec = userdef_spec)
# Extra info importing from a workspace
if(!extra_info)
return(result)
n_prespecified_out <- .jcall(jregression,"I","getPrespecifiedOutliersCount")
if(n_prespecified_out > 0 ){
result$userdef_spec$specification$outlier <- TRUE
outliers <- lapply(1:n_prespecified_out, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$Outlier;",
"getPrespecifiedOutlier",
as.integer(i-1))
})
type <- sapply(outliers, function(x) x$getCode())
date <- sapply(outliers, function(x) x$getPosition())
coeff <- sapply(outliers, function(x) x$getCoefficient())
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$outlier.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$outlier.coef <- TRUE
}
outlier_spec <- data.frame(type = type, date = date, coeff = coeff)
result$userdef_spec$outliers <- outlier_spec
}
n_userdefined_var <- .jcall(jregression,"I","getUserDefinedVariablesCount")
if(n_userdefined_var > 0 ){
result$userdef_spec$specification$variables <- TRUE
ud_vars <- lapply(1:n_userdefined_var, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$UserDefinedVariable;",
"getUserDefinedVariable",
as.integer(i-1))
})
type <- sapply(ud_vars, function(x) x$getComponent())
coeff <- sapply(ud_vars, function(x) x$getCoefficient())
var_names <- sapply(ud_vars, function(x) x$getName())
var_names_split <- strsplit(var_names,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
var_names <- base::make.names(var_names, unique = TRUE)
var_names <- gsub(".","_", var_names, fixed = TRUE)
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$variables.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$variables.coef <- TRUE
}
result$userdef_spec$variables$description <- data.frame(type = type,
coeff = coeff,
row.names = var_names)
# To check variables group names
# t <- context_dictionary$getTsVariableManagers()
# t$getNames()
#
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
#Calendar
user_td <- jtd$getUserVariables()
if(length(user_td) > 0 ){
var_names_split <- strsplit(user_td,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
var_names <- base::make.names(var_names, unique = TRUE)
var_names <- gsub(".","_", var_names, fixed = TRUE)
result$userdef_spec$specification$variables <-
TRUE
td_var_description <- data.frame(type = rep("Calendar",length(var_names)),
coeff = NA, row.names = var_names)
if(identical_na(result$userdef_spec$variables$description)){
result$userdef_spec$variables$description <- td_var_description
}else{
result$userdef_spec$variables$description <- rbind(result$userdef_spec$variables$description,
td_var_description)
}
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(!identical_na(result$userdef_spec$variables$series)){
var_series <- ts.union(result$userdef_spec$variables$series, var_series)
}
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
## Ramp effects
core_regression <- jregression$getCore()$getRegression()
jramps <- core_regression$getRamps()
nb_ramps <- core_regression$getRampsCount()
if (nb_ramps > 0) {
var_names <- sapply(seq_len(nb_ramps), function(x) jramps[[x]]$getDescription())
result$userdef_spec$specification$variables <-
TRUE
coeff <- NA
if (core_regression$hasFixedCoefficients()) {
coeff <- sapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
ramp_name <- jramp$getName()
fixed_coeff <- core_regression$getFixedCoefficients(ramp_name)
if (is.null(fixed_coeff)) {
NA
}else{
fixed_coeff
}
})
result$userdef_spec$specification$variables.coef <- TRUE
}
td_var_description <- data.frame(type = rep("Series", length(var_names)),
coeff = coeff, row.names = var_names)
if (identical_na(result$userdef_spec$variables$description)) {
result$userdef_spec$variables$description <-
td_var_description
}else{
result$userdef_spec$variables$description <-
rbind(result$userdef_spec$variables$description,
td_var_description)
}
if (!is.na(freq) || !identical_na(result$userdef_spec$variables$series)) {
if (is.na(freq))
freq <- frequency(result$userdef_spec$variables$series)
if (!identical_na(result$userdef_spec$variables$series)) {
start_ts <- start(result$userdef_spec$variables$series)
end_ts <- end(result$userdef_spec$variables$series)
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start = start_ts, end = end_ts,
start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
} else {
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
}
ramp_series <- ts(simplify2array(ramp_series),
start = start(ramp_series[[1]]),
frequency = frequency(ramp_series[[1]]))
if (!identical_na(result$userdef_spec$variables$series)) {
ramp_series <- ts.union(result$userdef_spec$variables$series, ramp_series)
}
if (is.mts(ramp_series))
colnames(ramp_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- ramp_series
}
}
Phi <- jarima$getPhi()
BPhi <- jarima$getBPhi()
Theta <- jarima$getTheta()
BTheta <- jarima$getBTheta()
arima_coefficients_spec <-
rbind(arimaCoef_jd2r(Phi),
arimaCoef_jd2r(BPhi),
arimaCoef_jd2r(Theta),
arimaCoef_jd2r(BTheta))
if(!is.null(arima_coefficients_spec) &&
any(arima_coefficients_spec$Type != "Undefined")){
result$arima.coef <- TRUE
result$arima.coef.spec <- arima_coefficients_spec
}
result
}
spec_TRAMO_jd2r <- function(spec = NA, context_dictionary = NULL,
extra_info = FALSE, freq = NA){
#Estimate
preliminary.check <- spec$getBasic()$isPreliminaryCheck()
jestimate <- .jcall(spec,"Ljdr/spec/tramoseats/EstimateSpec;","getEstimate")
jest.span <- .jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
estimate.type <- .jcall(jest.span,"S","getType")
estimate.d0 <- .jcall(jest.span,"S","getD0")
estimate.d1 <- .jcall(jest.span,"S","getD1")
estimate.n0 <- .jcall(jest.span,"I","getN0")
estimate.n1 <- .jcall(jest.span,"I","getN1")
estimate.span <- jd_span(type= estimate.type,d0=estimate.d0,d1=estimate.d1,n0=estimate.n0,n1=estimate.n1)
if (estimate.type %in% c("All", "First", "Last", "Excluding")) {
estimate.d0 <- estimate.d1 <- NA
} else if (estimate.type=="From") {
estimate.d0 <- NA
} else if (estimate.type=="To") {
estimate.d1 <- NA
}
estimate.tol <-.jcall(jestimate ,"D","getTol")
estimate.eml <-.jcall(jestimate ,"Z","isEml")
estimate.urfinal <-.jcall(jestimate ,"D","getUbp")
#Transform
jtransform <-.jcall(spec,"Ljdr/spec/tramoseats/TransformSpec;","getTransform")
transform.function <-.jcall(jtransform,"S","getFunction")
transform.fct <-.jcall(jtransform,"D","getFct")
#Regression
jregression<-.jcall(spec,"Ljdr/spec/tramoseats/RegressionSpec;","getRegression")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/tramoseats/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/tramoseats/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/tramoseats/EasterSpec;","getEaster")
tradingdays.mauto <-.jcall(jtd,"S","getAutomatic")
tradingdays.pftd <-.jcall(jtd,"D","getPftd")
tradingdays.option <- .jcall(jtd,"S","getOption")
if(tradingdays.option != "UserDefined"){
tradingdays.option <- .jcall(jtd,"S","getTradingDays")
}
tradingdays.leapyear <-.jcall(jtd,"Z","getLeapYear")
tradingdays.stocktd <-.jcall(jtd,"I","getW")
tradingdays.test <-.jcall(jtd,"S","getRegressionTestType")
easter.type <- .jcall(jeaster,"S","getOption")
easter.julian <-.jcall(jeaster,"Z","isJulian")
easter.duration <-.jcall(jeaster,"I","getDuration")
easter.test <-.jcall(jeaster,"Z","isTest")
#Outlier
joutlier<-.jcall(spec,"Ljdr/spec/tramoseats/OutlierSpec;","getOutlier")
joutlier.span <-.jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
outlier.enabled <-.jcall(joutlier,"Z","isOutliersDetectionEnabled")
outlier.type <- .jcall(joutlier.span,"S","getType")
outlier.d0 <- .jcall(joutlier.span,"S","getD0")
outlier.d1 <- .jcall(joutlier.span,"S","getD1")
outlier.n0 <- .jcall(joutlier.span,"I","getN0")
outlier.n1 <- .jcall(joutlier.span,"I","getN1")
outlier.span <- jd_span(type= outlier.type,d0=outlier.d0,d1=outlier.d1,n0=outlier.n0,n1=outlier.n1)
if (outlier.type %in% c("All", "First", "Last", "Excluding")) {
outlier.d0 <- outlier.d1 <- NA
} else if (outlier.type=="From") {
outlier.d0 <- NA
} else if (outlier.type=="To") {
outlier.d1 <- NA
}
outlier.ao <-.jcall(joutlier,"Z","isAO")
outlier.tc <-.jcall(joutlier,"Z","isTC")
outlier.ls <-.jcall(joutlier,"Z","isLS")
outlier.so <-.jcall(joutlier,"Z","isSO")
outlier.usedefcv <-.jcall(joutlier,"Z","isAutoVa")
outlier.cv <-.jcall(joutlier,"D","getVa")
outlier.eml <-.jcall(joutlier,"Z","isEML")
outlier.tcrate <-.jcall(joutlier,"D","getTCRate")
#Arima
jarima<-.jcall(spec,"Ljdr/spec/tramoseats/ArimaSpec;","getArima")
automdl.enabled <-.jcall(jarima,"Z","isEnabled")
automdl.acceptdefault <-.jcall(jarima,"Z","isAcceptDefault")
automdl.cancel <-.jcall(jarima,"D","getCancel")
automdl.ub1 <-.jcall(jarima,"D","getUb1")
automdl.ub2 <-.jcall(jarima,"D","getUb2")
automdl.armalimit <-.jcall(jarima,"D","getTsig")
automdl.reducecv <-.jcall(jarima,"D","getPc")
automdl.ljungboxlimit <-.jcall(jarima,"D","getPcr")
automdl.compare <-.jcall(jarima,"Z","isAmiCompare")
arima.mu <-.jcall(jarima,"Z","isMean")
arima.p <-.jcall(jarima,"I","getP")
arima.d <-.jcall(jarima,"I","getD")
arima.q <-.jcall(jarima,"I","getQ")
arima.bp <-.jcall(jarima,"I","getBP")
arima.bd <-.jcall(jarima,"I","getBD")
arima.bq <-.jcall(jarima,"I","getBQ")
#span matrix
type<-c(estimate.type,outlier.type)
d0<-c(estimate.d0,outlier.d0)
d1<-c(estimate.d1,outlier.d1)
n0<-c(estimate.n0,outlier.n0)
n1<-c(estimate.n1,outlier.n1)
span<-data.frame(type=type,d0=d0,d1=d1,n0=n0,n1=n1)
rownames(span)<-c("estimate","outlier")
arima.coef <- FALSE
userdef_spec <-list(specification = data.frame(outlier = FALSE,
outlier.coef = FALSE,
variables = FALSE,
variables.coef = FALSE,
stringsAsFactors = FALSE),
outliers = NA, variables = list(series = NA, description = NA))
arima.coef.spec <- NA
result <- list(preliminary.check = preliminary.check,
estimate.type = estimate.type,estimate.d0 = estimate.d0,estimate.d1 = estimate.d1,estimate.n0 = estimate.n0,
estimate.n1 = estimate.n1,estimate.span = estimate.span,estimate.tol = estimate.tol,estimate.eml = estimate.eml,
estimate.urfinal = estimate.urfinal,transform.function = transform.function,transform.fct = transform.fct,
tradingdays.mauto = tradingdays.mauto,tradingdays.pftd = tradingdays.pftd,tradingdays.option = tradingdays.option,
tradingdays.leapyear = tradingdays.leapyear,tradingdays.stocktd = tradingdays.stocktd,
tradingdays.test = tradingdays.test,easter.type = easter.type,easter.julian = easter.julian,easter.duration = easter.duration,
easter.test = easter.test,outlier.enabled = outlier.enabled,outlier.type = outlier.type,outlier.d0 = outlier.d0,
outlier.d1 = outlier.d1,outlier.n0 = outlier.n0,outlier.n1 = outlier.n1,outlier.span = outlier.span,outlier.ao = outlier.ao,
outlier.tc = outlier.tc,outlier.ls = outlier.ls,outlier.so = outlier.so,outlier.usedefcv = outlier.usedefcv,
outlier.cv = outlier.cv,outlier.eml = outlier.eml,outlier.tcrate = outlier.tcrate,automdl.enabled = automdl.enabled,
automdl.acceptdefault = automdl.acceptdefault,automdl.cancel = automdl.cancel,automdl.ub1 = automdl.ub1,
automdl.ub2 = automdl.ub2,automdl.armalimit = automdl.armalimit,automdl.reducecv = automdl.reducecv,
automdl.ljungboxlimit = automdl.ljungboxlimit,automdl.compare = automdl.compare,arima.mu = arima.mu,arima.p = arima.p,
arima.d = arima.d,arima.q = arima.q,arima.bp = arima.bp,arima.bd = arima.bd,arima.bq = arima.bq, span = span,
arima.coef = arima.coef,arima.coef.spec = arima.coef.spec,
userdef_spec = userdef_spec)
# Extra info importing from a workspace
if(!extra_info)
return(result)
n_prespecified_out <- .jcall(jregression,"I","getPrespecifiedOutliersCount")
if(n_prespecified_out > 0 ){
# Outlier.coef is set to TRUE: if the coefficient isn't fixed, the
# coef value will be equal to 0 and the outlier will be estimated
result$userdef_spec$specification$outlier <- TRUE
outliers <- lapply(1:n_prespecified_out, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$Outlier;",
"getPrespecifiedOutlier",
as.integer(i-1))
})
type <- sapply(outliers, function(x) x$getCode())
date <- sapply(outliers, function(x) x$getPosition())
coeff <- sapply(outliers, function(x) x$getCoefficient())
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$outlier.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$outlier.coef <- TRUE
}
outlier_spec <- data.frame(type = type, date = date, coeff = coeff)
result$userdef_spec$outliers <- outlier_spec
}
n_userdefined_var <- .jcall(jregression,"I","getUserDefinedVariablesCount")
if(n_userdefined_var > 0 ){
# variables.coef is set to TRUE: if the coefficient isn't fixed, the
# coef value will be equal to 0 and the variable will be estimated
result$userdef_spec$specification$variables <- TRUE
ud_vars <- lapply(1:n_userdefined_var, function(i){
.jcall(jregression,
"Ljdr/spec/ts/Utility$UserDefinedVariable;",
"getUserDefinedVariable",
as.integer(i-1))
})
type <- sapply(ud_vars, function(x) x$getComponent())
coeff <- sapply(ud_vars, function(x) x$getCoefficient())
var_names <- sapply(ud_vars, function(x) x$getName())
var_names_split <- strsplit(var_names,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
if(all(coeff == 0)){ #All coefficients are equal to 0: they are not fixed
result$userdef_spec$specification$variables.coef <- FALSE
coeff <- coeff * NA
}else{
result$userdef_spec$specification$variables.coef <- TRUE
}
result$userdef_spec$variables$description <- data.frame(type = type,
coeff = coeff,
row.names = var_names)
# To check variables group names
# t <- context_dictionary$getTsVariableManagers()
# t$getNames()
#
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
#Calendar
user_td <- jtd$getUserVariables()
if(length(user_td) > 0 ){
var_names_split <- strsplit(user_td,"[.]")
var_names <- sapply(var_names_split, function(x) x[2])
result$userdef_spec$specification$variables <-
TRUE
coeff <- NA
# if (core_regression$hasFixedCoefficients()) {
# # coeff <- sapply(seq_len(nb_ramps), function(i){
# # jramp <- jramps[[i]]
# # ramp_name <- jramp$getName()
# # fixed_coeff <- core_regression$getFixedCoefficients(ramp_name)
# # if (is.null(fixed_coeff)) {
# # NA
# # }else{
# # fixed_coeff
# # }
# # })
# # result$userdef_spec$specification$variables.coef <- TRUE
# }
td_var_description <- data.frame(type = rep("Calendar",length(var_names)),
coeff = coeff, row.names = var_names)
if(identical_na(result$userdef_spec$variables$description)){
result$userdef_spec$variables$description <- td_var_description
}else{
result$userdef_spec$variables$description <- rbind(result$userdef_spec$variables$description,
td_var_description)
}
if(!is.null(context_dictionary)){
var_series <- lapply(var_names_split,function(names){
ts_variable <- context_dictionary$getTsVariable(names[1],
names[2])
ts_jd2r(ts_variable$getTsData())
})
var_series <- ts(simplify2array(var_series),
start = start(var_series[[1]]), frequency = frequency(var_series[[1]]))
if(!identical_na(result$userdef_spec$variables$series)){
var_series <- ts.union(result$userdef_spec$variables$series, var_series)
}
if(is.mts(var_series))
colnames(var_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- var_series
}
}
## Ramp effects
core_regression <- jregression$getCore()$getRegression()
jramps <- core_regression$getRamps()
nb_ramps <- core_regression$getRampsCount()
if (nb_ramps > 0) {
var_names <- sapply(seq_len(nb_ramps), function(x) jramps[[x]]$getDescription())
result$userdef_spec$specification$variables <-
TRUE
coeff <- NA
if (core_regression$hasFixedCoefficients()) {
coeff <- sapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
ramp_name <- jramp$getName()
fixed_coeff <- core_regression$getFixedCoefficients(ramp_name)
if (is.null(fixed_coeff)) {
NA
}else{
fixed_coeff
}
})
result$userdef_spec$specification$variables.coef <- TRUE
}
td_var_description <- data.frame(type = rep("Series", length(var_names)),
coeff = coeff, row.names = var_names)
if (identical_na(result$userdef_spec$variables$description)) {
result$userdef_spec$variables$description <-
td_var_description
}else{
result$userdef_spec$variables$description <-
rbind(result$userdef_spec$variables$description,
td_var_description)
}
if (!is.na(freq) || !identical_na(result$userdef_spec$variables$series)) {
if (is.na(freq))
freq <- frequency(result$userdef_spec$variables$series)
if (!identical_na(result$userdef_spec$variables$series)) {
start_ts <- start(result$userdef_spec$variables$series)
end_ts <- end(result$userdef_spec$variables$series)
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start = start_ts, end = end_ts,
start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
} else {
ramp_series <- lapply(seq_len(nb_ramps), function(i){
jramp <- jramps[[i]]
jstart <- jramp$getStart()
jend <- jramp$getEnd()
start_ramp <- c(jstart$getYear(), jstart$getMonth())
end_ramp <- c(jend$getYear(), jend$getMonth())
ramp(start_ramp = start_ramp, end_ramp = end_ramp,
frequency = freq)
})
}
ramp_series <- ts(simplify2array(ramp_series),
start = start(ramp_series[[1]]),
frequency = frequency(ramp_series[[1]]))
if (!identical_na(result$userdef_spec$variables$series)) {
ramp_series <- ts.union(result$userdef_spec$variables$series, ramp_series)
}
if (is.mts(ramp_series))
colnames(ramp_series) <- rownames(result$userdef_spec$variables$description)
result$userdef_spec$variables$series <- ramp_series
}
}
#Arima
Phi <- jarima$getPhi()
BPhi <- jarima$getBPhi()
Theta <- jarima$getTheta()
BTheta <- jarima$getBTheta()
arima_coefficients_spec <-
rbind(arimaCoef_jd2r(Phi),
arimaCoef_jd2r(BPhi),
arimaCoef_jd2r(Theta),
arimaCoef_jd2r(BTheta))
if(!is.null(arima_coefficients_spec) &&
any(arima_coefficients_spec$Type != "Undefined")){
result$arima.coef <- TRUE
result$arima.coef.spec <- arima_coefficients_spec
}
result
}
specX11_jd2r <- function(spec = NA, freq = NA){
jx11 <- .jcall(spec,"Ljdr/spec/x13/X11Spec;","getX11")
if (!is.na(freq)) {
.jcall(jx11,"V","setFreq", as.integer(freq))
fullseasonalma <- .jcall(jx11,"[S","getFullSeasonalMA")
if (!is.null(fullseasonalma) && length(unique(fullseasonalma)) > 1) {
seasonalma <- paste(fullseasonalma, collapse = ", ")
}else{
seasonalma <- .jcall(jx11,"S","getSeasonalMA")
}
}else{
seasonalma <- .jcall(jx11,"S","getSeasonalMA")
}
mode <- .jcall(jx11,"S","getMode")
seasonalComp <- .jcall(jx11,"Z","isSeasonal")
lsigma <- .jcall(jx11,"D","getLSigma")
usigma <- .jcall(jx11,"D","getUSigma")
trendAuto <- .jcall(jx11,"Z","isAutoTrendMA")
trendma <- .jcall(jx11,"I","getTrendMA")
fcasts <- .jcall(jx11,"I","getForecastHorizon")
bcasts <- .jcall(jx11,"I","getBackcastHorizon")
excludeFcasts <- .jcall(jx11,"Z","isExcludefcst")
calendarSigma <- jx11$getCalendarSigma()$toString()
sigmaVector <- .jcall(jx11,returnSig = "[S", "getSigmavec") #TODO
if (length(sigmaVector) == 0) {
sigmaVector <- NA
}else{
sigmaVector <- paste(sigmaVector, collapse = ", ")
}
var <- data.frame(mode,seasonalComp,lsigma,usigma,trendAuto,trendma,seasonalma,
fcasts, bcasts,
calendarSigma, sigmaVector,
excludeFcasts,
stringsAsFactors = FALSE)
names(var) <- sprintf("x11.%s",
c("mode","seasonalComp","lsigma","usigma",
"trendAuto","trendma","seasonalma",
"fcasts","bcasts",
"calendarSigma", "sigmaVector",
"excludeFcasts"))
return(var)
}
specSeats_jd2r <- function(spec = NA){
jseats <- .jcall(spec,"Ljdr/spec/tramoseats/SeatsSpec;","getSeats")
predictionLength <- .jcall(jseats,"I","getPredictionLength")
approx <- .jcall(jseats,"S","getApproximationMode")
maBoundary <- .jcall(jseats,"D","getXl")
trendBoundary <- .jcall(jseats,"D","getRMod")
seasdBoundary <- .jcall(jseats,"D","getSMod")
seasdBoundary1 <- .jcall(jseats,"D","getSMod1")
seasTol <- .jcall(jseats,"D","getEpsPhi")
method <- .jcall(jseats,"S","getMethod")
var <- list(predictionLength, approx,maBoundary,trendBoundary,seasdBoundary,seasdBoundary1,seasTol,
method)
names(var) <- c("predictionLength", "approx", "maBoundary", "trendBoundary",
"seasdBoundary","seasdBoundary1","seasTol",
"method")
return(var)
}
# Functions to introduce modifications in the java object
span_r2jd <- function(jsobjct = NA, type = NA, d0=NA, d1=NA, n0 = NA, n1=NA){
if (type =="All") {
.jcall(jsobjct,"V","all")
} else if (type=="From"){
.jcall(jsobjct,"V","from",d0)
} else if (type=="To"){
.jcall(jsobjct,"V","to",d1)
} else if (type=="Between"){
.jcall(jsobjct,"V","between",d0,d1)
} else if (type=="First"){
.jcall(jsobjct,"V","first",n0)
} else if (type=="Last"){
.jcall(jsobjct,"V","last",n1)
} else if (type=="Excluding"){
.jcall(jsobjct,"V","excluding",n0,n1)
}
}
preOut_r2jd <- function(jsobjct = NA, coefEna = NA, out = NA, outDate = NA, outCoef = NA){
coef <- if (coefEna == TRUE) {outCoef} else {rep(0,length(out))}
for (i in 1:length(out)){
.jcall(jsobjct,"V","addPrespecifiedOutlier", out[i],outDate[i],coef[i])
}
}
preVar_r2jd <- function(jsobjct = NA, jsdict = NA, coefEna = NA,
prevar_spec = list(series = NA, description = data.frame(NA,NA)),
jtd = NA) {
series = prevar_spec$series
varType = prevar_spec$description[,1]
varCoef = prevar_spec$description[,2]
varNames <- rownames(prevar_spec$description)
nvar <- if (is.mts(series)) {dim(series)[2]} else if (is.ts(series)) {1} else {0}
type <- if (all(is.na(varType))) {rep("Undefined",nvar)} else {as.character(varType)}
coef <- if (coefEna == TRUE) {varCoef} else {rep(0,nvar)}
if(nvar == 0)
return(NULL)
calendar_def <- grep("Calendar",type)
n_calendar_def <- length(calendar_def)
jcoreg = jsobjct$getCore()$getRegression()
if (nvar == 1){
if(n_calendar_def == 1){
.jcall(jsdict,"V","add",varNames,ts_r2jd(series))
if(coef!=0){
jcoreg$setFixedCoefficients(paste0("td|r@",varNames),
.jarray(coef))
}
.jcall(jtd,"V","setUserVariables", .jarray(paste0("r.",varNames)))
}else{
.jcall(jsdict,"V","add",varNames,ts_r2jd(series))
.jcall(jsobjct,"V","addUserDefinedVariable",varNames,type[1],coef[1])
}
}else{
if(n_calendar_def == 0 | n_calendar_def == nvar){
if(n_calendar_def >0){
for (i in 1:nvar){
.jcall(jsdict,"V","add",varNames[i],ts_r2jd(series[,i]))
if(coef[i]!=0){
jcoreg$setFixedCoefficients(paste0("td|r@",varNames[i]),
.jarray(coef[i]))
}
}
.jcall(jtd,"V","setUserVariables",
paste0("r.",varNames)
)
}else{
for (i in 1:nvar){
.jcall(jsdict,"V","add",varNames[i],ts_r2jd(series[,i]))
.jcall(jsobjct,"V","addUserDefinedVariable",varNames[i],type[i],coef[i])
}
}
}else{
i_ud <- (1:nvar)[-calendar_def]
for (i in i_ud){
.jcall(jsdict,"V","add", varNames[i],
ts_r2jd(series[, i]))
.jcall(jsobjct,"V","addUserDefinedVariable", varNames[i],
type[i], coef[i])
}
for (i in calendar_def){
.jcall(jsdict,"V","add",varNames[i],
ts_r2jd(series[, i]))
if(coef[i]!=0){
jcoreg$setFixedCoefficients(paste0("td|r@",varNames[i]),
.jarray(coef[i]))
}
}
.jcall(jtd,"V","setUserVariables",
.jarray(paste0("r.",varNames[calendar_def])))
}
}
}
arimaCoef_r2jd <- function(jsobjct = NA, acoef = NA, p = NA , q = NA, bp = NA, bq = NA){
typ <- acoef[,1]
val <- acoef[,2]
np <- length(typ)
par <- val
par[is.na(val) | val==0 | typ=="Undefined"] <- NaN
fix<-rep(FALSE,np)
fix[typ=="Fixed" & !is.na(par)] <- TRUE
n <-1
if (p!=0){
param <- parameters_r2jd(par[n:(n+p-1)],fix[n:(n+p-1)])
.jcall(jsobjct,"V","setPhi",param)
n <- n+p
}
if (q!=0){
param <- parameters_r2jd(par[n:(n+q-1)],fix[n:(n+q-1)])
.jcall(jsobjct,"V","setTheta",param)
n <- n+q
}
if (bp!=0){
param <- parameters_r2jd(par[n:(n+bp-1)],fix[n:(n+bp-1)])
.jcall(jsobjct,"V","setBPhi",param)
n <- n+bp
}
if (bq!=0){
param <- parameters_r2jd(par[n:(n+bq-1)],fix[n:(n+bq-1)])
.jcall(jsobjct,"V","setBTheta",param)
}
}
arimaCoef_jd2r <- function(jparams){
if (is.jnull(jparams))
return(NULL)
param<-.jcastToArray(jparams)
len <- length(param)
if (len==0)
return (NULL)
param_name <- deparse(substitute(jparams))
Type <- sapply(param, function(x) x$getType()$toString())
Value <- sapply(param, function(x) x$getValue())
data_param <- data.frame(Type = Type, Value = Value)
rownames(data_param) <- sprintf("%s(%i)",
param_name,
1:len)
data_param
}
spec_regarima_X13_r2jd <- function(spec = NA, jdspec = NA){
if (is.null(s_estimate(spec)))
return(.jnew("jdr/spec/ts/Utility$Dictionary"))
est <- s_estimate(spec)
trans <- s_transform(spec)
usrspc <- s_usrdef(spec)
outF <- s_preOut(spec)
varF <- s_preVar(spec)
td <- s_td(spec)
easter <- s_easter(spec)
outliers <- s_out(spec)
arimaspc <- s_arima(spec)
arimacoF <- s_arimaCoef(spec)
span <- s_span(spec)
#Estimate
jestimate <-.jcall(jdspec,"Ljdr/spec/x13/EstimateSpec;","getEstimate")
jest.span <-.jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = jest.span, type = span[1,1], d0=as.character(span[1,2]), d1=as.character(span[1,3]),
n0 = as.integer(span[1,4]), n1=as.integer(span[1,5]))
.jcall(jestimate ,"V","setTol", as.numeric(est["tolerance"]))
jdspec$getBasic()$setPreliminaryCheck(est[1, "preliminary.check"])
#Transform
jtransform <-.jcall(jdspec,"Ljdr/spec/x13/TransformSpec;","getTransform")
.jcall(jtransform,"V","setFunction",as.character(trans[1]))
.jcall(jtransform,"V","setAdjust",as.character(trans[2]))
.jcall(jtransform,"V","setAic",as.numeric(trans[3]))
#Regression
jregression<-.jcall(jdspec,"Ljdr/spec/x13/RegressionSpec;","getRegression")
#Pre-specified outliers
.jcall(jregression,"V","clearPrespecifiedOutliers")
if (usrspc[1]==TRUE)
preOut_r2jd(jsobjct = jregression, coefEna= usrspc[2], out= as.character(outF[,1]),outDate= as.character(outF[,2]),outCoef= outF[,3])
#User-defined variables
.jcall(jregression,"V","clearUserDefinedVariables")
jdictionary <- .jnew("jdr/spec/ts/Utility$Dictionary")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/x13/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/x13/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/x13/EasterSpec;","getEaster")
.jcall(jeaster,"V","setTest", as.character(easter[4]))
.jcall(jeaster,"V","setJulian",as.logical(easter[2]))
.jcall(jeaster,"V","setDuration", as.integer(easter[3]))
.jcall(jeaster,"V","setEnabled",as.logical(easter[1]))
#Calendar options:
if(td[1] == "UserDefined"){
.jcall(jtd,"V","setOption","UserDefined")
}else{
if (td[4]==0) {
.jcall(jtd,"V","setW",as.integer(td[4]))
.jcall(jtd,"V","setTradingDays",as.character(td[1]))
}else{
.jcall(jtd,"V","setTradingDays",as.character(td[1]))
.jcall(jtd,"V","setW",as.integer(td[4]))
}
.jcall(jtd,"V","setAutoAdjust",as.logical(td[2]))
.jcall(jtd,"V","setLengthOfPeriod", as.character(td[3]))
}
.jcall(jtd,"V","setTest",as.character(td[5]))
#user-defined specification
if (usrspc[3]==TRUE)
preVar_r2jd(jsobjct = jregression, jsdict = jdictionary, coefEna = usrspc[4],
prevar_spec = varF, jtd = jtd)
#Outlier
joutlier<-.jcall(jdspec,"Ljdr/spec/x13/OutlierSpec;","getOutliers")
joutlier.span <-.jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = joutlier.span, type = span[2,1], d0=as.character(span[2,2]), d1=as.character(span[2,3]),
n0 = as.integer(span[2,4]), n1=as.integer(span[2,5]))
.jcall(joutlier,"V","setAO",as.logical(outliers[3]))
.jcall(joutlier,"V","setTC",as.logical(outliers[4]))
.jcall(joutlier,"V","setLS",as.logical(outliers[5]))
.jcall(joutlier,"V","setSO",as.logical(outliers[6]))
# Default critical value
if (outliers[7] == TRUE){
.jcall(joutlier,"V","setDefaultVa",as.logical(outliers[7]))
}else{
.jcall(joutlier,"V","setVa",as.numeric(outliers[8]))
}
.jcall(joutlier,"V","setMethod",as.character(outliers[9]))
.jcall(joutlier,"V","setTCRate",as.numeric(outliers[10]))
.jcall(joutlier,"V","setEnabled",as.logical(outliers[1]))
#Arima
jarima<-.jcall(jdspec,"Ljdr/spec/x13/ArimaSpec;","getArima")
.jcall(jarima,"V","setAmiEnabled",as.logical(arimaspc[1]))
.jcall(jarima,"V","setAcceptDefault",as.logical(arimaspc[2]))
.jcall(jarima,"V","setCancel",as.numeric(arimaspc[3]))
.jcall(jarima,"V","setUb1", as.numeric(arimaspc[4]))
.jcall(jarima,"V","setUb2", as.numeric(arimaspc[5]))
.jcall(jarima,"V","setMixed", as.logical(arimaspc[6]))
.jcall(jarima,"V","setBalanced", as.logical(arimaspc[7]))
.jcall(jarima,"V","setTsig",as.numeric(arimaspc[8]))
.jcall(jarima,"V","setPredCV",as.numeric(arimaspc[9]))
.jcall(jarima,"V","setPcr", as.numeric(arimaspc[10]))
.jcall(jarima,"V","setUbFinal", as.numeric(arimaspc[11]))
.jcall(jarima,"V","setMean",as.logical(arimaspc[12]))
.jcall(jarima,"V","setP",as.integer(arimaspc[13]))
.jcall(jarima,"V","setD",as.integer(arimaspc[14]))
.jcall(jarima,"V","setQ",as.integer(arimaspc[15]))
.jcall(jarima,"V","setBP",as.integer(arimaspc[16]))
.jcall(jarima,"V","setBD",as.integer(arimaspc[17]))
.jcall(jarima,"V","setBQ",as.integer(arimaspc[18]))
# Fixed ARIMA coefficients
if (arimaspc[19]==TRUE)
arimaCoef_r2jd(jsobjct = jarima, acoef = arimacoF, p = as.numeric(arimaspc[13]) , q = as.numeric(arimaspc[15]),
bp = as.numeric(arimaspc[16]), bq = as.numeric(arimaspc[18]))
return(jdictionary)
}
spec_TRAMO_r2jd <- function(spec = NA, jdspec =NA){
est <- s_estimate(spec)
trans <- s_transform(spec)
usrspc <- s_usrdef(spec)
outF <- s_preOut(spec)
varF <- s_preVar(spec)
td <- s_td(spec)
easter <- s_easter(spec)
outliers <- s_out(spec)
arimaspc <- s_arima(spec)
arimacoF <- s_arimaCoef(spec)
span <- s_span(spec)
#Estimate
jdspec$getBasic()$setPreliminaryCheck(est[1, "preliminary.check"])
jestimate <-.jcall(jdspec,"Ljdr/spec/tramoseats/EstimateSpec;","getEstimate")
jest.span <-.jcall(jestimate,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = jest.span, type = span[1,1], d0=as.character(span[1,2]), d1=as.character(span[1,3]),
n0 = as.integer(span[1,4]), n1=as.integer(span[1,5]))
.jcall(jestimate ,"V","setTol",as.numeric(est["tolerance"]))
.jcall(jestimate ,"V","setEml",as.logical(est["exact_ml"]))
.jcall(jestimate ,"V","setUbp",as.numeric(est["urfinal"]))
#Transform
jtransform <-.jcall(jdspec,"Ljdr/spec/tramoseats/TransformSpec;","getTransform")
.jcall(jtransform,"V","setFunction",as.character(trans[1]))
.jcall(jtransform,"V","setFct", as.numeric(trans[2]))
#Regression
jregression<-.jcall(jdspec,"Ljdr/spec/tramoseats/RegressionSpec;","getRegression")
#Pre-specified outliers
.jcall(jregression,"V","clearPrespecifiedOutliers")
if (usrspc[1]==TRUE)
preOut_r2jd(jsobjct = jregression, coefEna= usrspc[2],out= as.character(outF[,1]), outDate= as.character(outF[,2]),outCoef= outF[,3])
#User-defined variables
.jcall(jregression,"V","clearUserDefinedVariables")
jdictionary <- .jnew("jdr/spec/ts/Utility$Dictionary")
#Calendar
jcalendar<-.jcall(jregression,"Ljdr/spec/tramoseats/CalendarSpec;","getCalendar")
jtd<-.jcall(jcalendar,"Ljdr/spec/tramoseats/TradingDaysSpec;","getTradingDays")
jeaster<-.jcall(jcalendar,"Ljdr/spec/tramoseats/EasterSpec;","getEaster")
.jcall(jtd,"V","setAutomatic", as.character(td[1]))
.jcall(jtd,"V","setPftd", as.numeric(td[2]))
if(td[3] == "UserDefined"){
.jcall(jtd,"V","setOption","UserDefined")
}else{
if (td[5]==0) {
# .jcall(jtd,"V","setW",as.integer(td[5]))
.jcall(jtd,"V","setTradingDays",as.character(td[3]))
}else{
.jcall(jtd,"V","setTradingDays",as.character(td[3]))
.jcall(jtd,"V","setW",as.integer(td[5]))
}
.jcall(jtd,"V","setLeapYear",as.logical(td[4]))
}
.jcall(jtd,"V","setRegressionTestType", as.character(td[6]))
.jcall(jeaster,"V","setOption",as.character(easter[1]))
.jcall(jeaster,"V","setJulian",as.logical(easter[2]))
.jcall(jeaster,"V","setDuration",as.integer(easter[3]))
.jcall(jeaster,"V","setTest",as.logical(easter[4]))
#user-defined specification
if (usrspc[3]==TRUE)
preVar_r2jd(jsobjct = jregression, jsdict = jdictionary, coefEna = usrspc[4],
prevar_spec = varF, jtd = jtd)
#Outlier
joutlier<-.jcall(jdspec,"Ljdr/spec/tramoseats/OutlierSpec;","getOutlier")
joutlier.span <-.jcall(joutlier,"Ljdr/spec/ts/SpanSelector;","getSpan")
span_r2jd(jsobjct = joutlier.span, type = span[2,1], d0=as.character(span[2,2]), d1=as.character(span[2,3]),
n0 = as.integer(span[2,4]), n1=as.integer(span[2,5]))
.jcall(joutlier,"V","setOutliersDetectionEnabled", as.logical(outliers[1]))
.jcall(joutlier,"V","setAO", as.logical(outliers[3]))
.jcall(joutlier,"V","setTC", as.logical(outliers[4]))
.jcall(joutlier,"V","setLS", as.logical(outliers[5]))
.jcall(joutlier,"V","setSO", as.logical(outliers[6]))
# Default critical value
if (outliers[7] == TRUE){
.jcall(joutlier,"V","setAutoVa", as.logical(outliers[7]))
}else{
.jcall(joutlier,"V","setVa", as.numeric(outliers[8]))
}
.jcall(joutlier,"V","setEML", as.logical(outliers[9]))
.jcall(joutlier,"V","setTCRate", as.numeric(outliers[10]))
#Arima
jarima<-.jcall(jdspec,"Ljdr/spec/tramoseats/ArimaSpec;","getArima")
.jcall(jarima,"V","setEnabled", as.logical(arimaspc[1]))
.jcall(jarima,"V","setAcceptDefault", as.logical(arimaspc[2]))
.jcall(jarima,"V","setCancel", as.numeric(arimaspc[3]))
.jcall(jarima,"V","setUb1", as.numeric(arimaspc[4]))
.jcall(jarima,"V","setUb2", as.numeric(arimaspc[5]))
.jcall(jarima,"V","setTsig", as.numeric(arimaspc[6]))
.jcall(jarima,"V","setPc", as.numeric(arimaspc[7]))
.jcall(jarima,"V","setPcr", as.numeric(arimaspc[8]))
.jcall(jarima,"V","setAmiCompare", as.logical(arimaspc[9]))
.jcall(jarima,"V","setMean", as.logical(arimaspc[10]))
.jcall(jarima,"V","setP", as.integer(arimaspc[11]))
.jcall(jarima,"V","setD", as.integer(arimaspc[12]))
.jcall(jarima,"V","setQ", as.integer(arimaspc[13]))
.jcall(jarima,"V","setBP", as.integer(arimaspc[14]))
.jcall(jarima,"V","setBD", as.integer(arimaspc[15]))
.jcall(jarima,"V","setBQ", as.integer(arimaspc[16]))
# Fixed ARIMA coefficients
if (arimaspc[17]==TRUE)
arimaCoef_r2jd(jsobjct = jarima, acoef = arimacoF, p = as.numeric(arimaspc[11]) , q = as.numeric(arimaspc[13]),
bp = as.numeric(arimaspc[14]), bq = as.numeric(arimaspc[16]))
return(jdictionary)
}
specX11_r2jd <- function(rspec = NA, jdspec = NA , freq = NA){
x11 <- s_x11(rspec)
jx11 <- .jcall(jdspec,"Ljdr/spec/x13/X11Spec;","getX11")
seasonalma <- unlist(strsplit(as.character(x11[["x11.seasonalma"]]),
split = ", "))
len.ma <- length(seasonalma)
.jcall(jx11,"V","setMode",as.character(x11[["x11.mode"]]))
.jcall(jx11,"V","setSeasonal",as.logical(x11[["x11.seasonalComp"]]))
.jcall(jx11,"V","setLSigma",as.numeric(x11[["x11.lsigma"]]))
.jcall(jx11,"V","setUSigma",as.numeric(x11[["x11.usigma"]]))
if (x11[["x11.trendAuto"]]) {
.jcall(jx11,"V","setAutoTrendMA",as.logical(TRUE))
}else{
.jcall(jx11,"V","setAutoTrendMA",as.logical(FALSE))
.jcall(jx11,"V","setTrendMA",as.integer(x11[["x11.trendma"]]))
}
.jcall(jx11,"V","setFreq", as.integer(freq))
if (len.ma == 1) {
.jcall(jx11, "V", "setSeasonalMA", seasonalma)
seasma <- seasonalma
}else if (len.ma != freq) {
.jcall(jx11,"V","setSeasonalMA","Msr")
warning(paste0("wrong frequency of the x11.seasonalma (",
len.ma, " instead of ", freq, ").",
"\nPre-specified seasonal filters will be ignored (x11.seasonalma=\"Msr\")."),
call. = FALSE)
seasma <- "Msr"
} else {
.jcall(jx11,"V","setFullSeasonalMA",seasonalma)
seasma <- as.character(x11[["x11.seasonalma"]])
}
.jcall(jx11, "V", "setForecastHorizon", as.integer(x11[["x11.fcasts"]]))
.jcall(jx11, "V", "setBackcastHorizon", as.integer(x11[["x11.bcasts"]]))
.jcall(jx11, "V", "setCalendarSigma", x11[["x11.calendarSigma"]])
if (x11[["x11.calendarSigma"]] == "Select" &&
!identical_na(x11[["x11.sigmaVector"]])) {
# sigmaVector is change only if x11.calendarSigma is set to "Select"
sigmaVector <- unlist(strsplit(as.character(x11[["x11.sigmaVector"]]),
split = ", "))
if (length(sigmaVector) != freq) {
warning(paste0("Wrong frequency of the x11.sigmaVector (",
length(sigmaVector), " instead of ", freq, ").",
"\nThis parameter will be ignored."), call. = FALSE)
} else {
.jcall(jx11, "V", "setSigmavec", sigmaVector)
}
}
.jcall(jx11, "V", "setExcludefcst", as.logical(x11[["x11.excludeFcasts"]]))
return(seasma)
}
specSeats_r2jd <- function(rspec = NA, jdspec = NA){
seats <- s_seats(rspec)
jseats <-.jcall(jdspec,"Ljdr/spec/tramoseats/SeatsSpec;","getSeats")
.jcall(jseats,"V","setPredictionLength", as.integer(seats[["seats.predictionLength"]]))
.jcall(jseats,"V","setApproximationMode", as.character(seats[["seats.approx"]]))
.jcall(jseats,"V","setXl", as.numeric(seats[["seats.maBoundary"]]))
.jcall(jseats,"V","setRMod", as.numeric(seats[["seats.trendBoundary"]]))
.jcall(jseats,"V","setSMod", as.numeric(seats[["seats.seasdBoundary"]]))
.jcall(jseats,"V","setSMod1", as.numeric(seats[["seats.seasdBoundary1"]]))
.jcall(jseats,"V","setEpsPhi",as.numeric(seats[["seats.seasTol"]]))
.jcall(jseats,"V","setMethod", as.character(seats[["seats.method"]]))
}
Try the RJDemetra 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.