search

SMAC-Group / gmwm

Generalized Method of Wavelet Moments (GMWM) is an estimation technique for the parameters of time series models. It uses the wavelet variance in a moment matching approach that makes it particularly suitable for the estimation of certain state-space models.
Other
28 stars 14 forks source link

Add `MA1()` and `ARMA11()` ts.model objects #161

Closed coatless closed 8 years ago

coatless commented 8 years ago

Under the current setup, both MA1() and ARMA11() objects would be able to work within the model selection paradigm.

Only the following would not work (numerical derivative issues): MA(q) AR(p) ARMA(p,q) SARMA(p,q,P,Q,s) SARIMA(p,d,q,P,D,Q,s) ARIMA(p,d,q)

coatless commented 8 years ago

Fixes arma11_to_wv() form

fb48bf1860bd681fcea550d3ba2acbab572db97c (unstable for phi = 0)

89a00a470b453ed3a39e4c01eb74c3c29479a7f9 (stable for all values)

coatless commented 8 years ago

@Wenchao-Yang

For the earlier graphs, was checking for a duplicate AR1 process hard coded? If that is the case, it needs to be switched to using addSpaceIfDuplicate.

coatless commented 8 years ago

The addSpaceIfDuplicate was taken care of in PR #171

coatless commented 8 years ago

The only thing left really is to address guessing for MA(1) and ARMA(1,1). They are currently using the old methods... I'll leave that for #162