predict.tspca: Make predictions from a '"tspca"' object

In HDTSA: High Dimensional Time Series Analysis Tools

Make predictions from a "tspca" object

Description

This function makes predictions from a "tspca" object.

Usage

## S3 method for class 'tspca'
predict(
  object,
  newdata = NULL,
  n.ahead = 10,
  control_ARIMA = list(),
  control_VAR = list(),
  ...
)

Arguments

object	An object of class "tspca" constructed by PCA_TS.
newdata	Optional. A new data matrix to predict from.
n.ahead	An integer specifying the number of steps ahead for prediction.
control_ARIMA	A list of arguments passed to the function auto.arima() of forecast. See 'Details' and the manual of auto.arima(). The default is list(max.d = 0, max.q = 0, ic = "aic").
control_VAR	A list of arguments passed to the function VAR() of vars. See 'Details' and the manual of VAR(). The default is list(type = "const", lag.max = 6, ic = "AIC").
...	Currently not used.

Details

Having obtained \hat{\bf x}_t using the PCA_TS function, which is segmented into q uncorrelated subseries \hat{\bf x}_t^{(1)},\ldots,\hat{\bf x}_t^{(q)}, the forecasting for {\bf y}_t can be performed in two steps:

Step 1. Get the h-step ahead forecast \hat{\bf x}_{n+h}^{(j)} (j=1,\ldots,q) by using a VAR model (if the dimension of \hat{\bf x}_t^{(j)} is larger than 1) or an ARIMA model (if the dimension of \hat{\bf x}_t^{(j)} is 1). The orders of VAR and ARIMA models are determined by AIC by default. Otherwise, they can also be specified by users through the arguments control_VAR and control_ARIMA, respectively.

Step 2. Let \hat{\bf x}_{n+h} = (\{\hat{\bf x}_{n+h}^{(1)}\}',\ldots,\{\hat{\bf x}_{n+h}^{(q)}\}')'. The forecasted value for {\bf y}_t is obtained by \hat{\bf y}_{n+h}= \hat{\bf B}^{-1}\hat{\bf x}_{n+h}.

Value

Y_pred

A matrix of predicted values.

See Also

PCA_TS

Examples

library(HDTSA)
data(FamaFrench, package = "HDTSA")

## Remove the market effects
reg <- lm(as.matrix(FamaFrench[, -c(1:2)]) ~ as.matrix(FamaFrench$MKT.RF))
Y_2d = reg$residuals

res_pca <- PCA_TS(Y_2d, lag.k = 5, thresh = TRUE)
pred_pca_Y <- predict(res_pca, n.ahead = 1)

HDTSA documentation built on April 3, 2025, 11:07 p.m.