Adds constant mean shifts to a multivariate time series by applying a fixed jump vector at evenly spaced change points. After each change point, all subsequent observations are shifted by the specified vector.
add_jumps(data, delta, num)A numeric matrix of the same dimension as `data`, containing the adjusted series with added mean shifts.
Numeric matrix of dimension \(n \times p\), representing the time series data.
Numeric vector of length \(p\), specifying the shift magnitudes added to each variable after each change point.
Integer; number of change points. The data are divided evenly
into num + 1 segments, and delta is added cumulatively after
each change point.
The total length of the time series is denoted by \(n\). Change points are
placed at evenly spaced locations given by
\(k \lfloor n / (num + 1) \rfloor\), for \(k = 1, \ldots, num\). After each
change point, a constant shift vector delta is added to all subsequent
observations. This construction produces synthetic data with known and
controlled mean shifts, making the function useful for simulation studies and
benchmarking change point detection methods.