View source: R/sigex.adhocextract.r
sigex.adhocextract | R Documentation |
Background: A sigex model consists of process x = sum y, for stochastic components y. Each component process y_t is either stationary or is reduced to stationarity by application of a differencing polynomial delta(B), i.e. w_t = delta(B) y_t is stationary. We have a model for each w_t process, and can compute its autocovariance function (acf), and denote its autocovariance generating function (acgf) via gamma_w (B). The signal extraction filter for y_t is determined from this acgf and delta. param is the name for the model parameters entered into a list object with a more intuitive structure, whereas psi refers to a vector of real numbers containing all hyper-parameters (i.e., reals mapped bijectively to the parameter manifold) together with imaginary component flagging whether the hyper-parameter is fixed for purposes of estimation.
sigex.adhocextract(psi, mdl, data.ts, adhoc, shift, horizon, needMSE)
psi |
A vector of all the real hyper-parameters |
mdl |
The specified sigex model, a list object |
data.ts |
A T x N matrix ts object; any values to be imputed must be encoded with NA in that entry. The NA is for missing value, or an enforced imputation (e.g. extreme-value adjustment). |
adhoc |
An array N x N x L, where L is length |
shift |
Gives the integer offset for the adhoc filter: filter coefficients have indices -shift,...,0,...,L-1-shift set shift = 0 for a causal filter |
horizon |
A non-negative integer indicating how many forecasts and aftcasts of the signal should be generated |
needMSE |
A binary flag, set to 1 if you want MSE based on casting error, or if there are any missing values; else (with value 0) the routine runs faster and returns zero for the MSE. |
Notes: method does midcasts for specified time indices, applies ad hoc filter for signal (given by adhoc), of length L and offset "shift". generates signal at all time points t in seq(L-shift-horizon,T-shift+horizon).
list object of extract.sig, upp, and low extract.sig: (T+H) x N matrix of the signal estimates, where H is twice the length of horizon upp: as extract.sig, plus twice the standard error low: as extract.sig, minus twice the standard error
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.