matrix2Tensor3D: matrix2Tensor3D

In jdestefani/ExtendedDFML: Extended Dynamic Factor Machine Learner

Description

Auxiliary function to embed a 2D matrix into a 3D tensor.

Usage

matrix2Tensor3D(X, time_window, shift = 1, padNA = FALSE)

Arguments

X	Input matrix
time_window	Integer specifying the size of the time window (aka model order, 2nd dimension in tensor).
shift	Integer specifying how far the sliding window should be shifted in order to create the tensor (N.B. shift < time_window implies overlapping data)
padNA	Boolean specifying whether NAs should be used to pad tensors in case the input matrix size is not multiple of the time_window.

Value

3D Tensor (Samples,Time Window,Features) containing the embedded form of X (the second dimension is fixed to time_window)

Examples

X <- EuStockMarkets
time_window <- 5
X_windowed <- matrix2Tensor3D(X,time_window,time_window)

jdestefani/ExtendedDFML documentation built on Dec. 20, 2021, 10:04 p.m.