rescale.CPST | R Documentation |
This rescaling procedure is needed to cope with possible heteroscedasticity in the data. Simply it's achieved by taking a cumulative sum of the first difference normalized by the non-parametric local estimate of the variance.
rescale.CPST(d.y, x, deter, adf.lag, max.lag)
d.y |
A series of first differences. |
x |
A matrix of ADF variables. |
deter |
A matrix of deterministic variables for detrending. |
adf.lag |
A lag of the corresponding ADF model. |
max.lag |
The maximum possible lag. |
A rescaled series.
Cavaliere, Giuseppe, Peter C. B. Phillips, Stephan Smeekes, and A. M. Robert Taylor. “Lag Length Selection for Unit Root Tests in the Presence of Nonstationary Volatility.” Econometric Reviews 34, no. 4 (April 21, 2015): 512–36. https://doi.org/10.1080/07474938.2013.808065.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.