This thread is about the StatFact add-in for extracting static factors from macro-panels.
The add-in allows you to estimate factors from 'macro-panels' by the method of asymptotic principle components (as per Stock and Watson 2002). Taking a group object as input, it provides multiple options (through both the command line and a GUI) to estimate Bai and Ng (2002) information criteria (PC, IPC, AIC3 and BIC3), to take first differences of the whole group, to standardize the group, to output loadings, to normalize the factors, and to output graphics of the factors themselves, loadings, scree plots and % of variance explained.
StatFact
Moderators: EViews Gareth, EViews Moderator, EViews Esther
Re: StatFact
Dimensional reduction in large macroeconomic models are very important and this is one of the most commonly used methods. Nice to have it as an EViews add-in now. Thanks for the effort and sharing.
Re: StatFact
EDIT: I found the answer to my question by accident. You first need to select a group object of multiple series. Then the statfact option appears in the interface.
Hi guys,
I would like to use the statfact add-in in order to determine the appropriate number of factors to be estimated by principal components for a large marco data set. However, after downloading and installing the add-in, no GUI option appears in my EViews menu (version 10, university edition). Does anybody know if the add-in is still working by means of command:
?
If it does not work under my EViews version, do you know of a suitable alternative?
Thank you in advance.
All help is much appreciated!
Hi guys,
I would like to use the statfact add-in in order to determine the appropriate number of factors to be estimated by principal components for a large marco data set. However, after downloading and installing the add-in, no GUI option appears in my EViews menu (version 10, university edition). Does anybody know if the add-in is still working by means of command:
Code: Select all
group object.statfact(options) factornames #
If it does not work under my EViews version, do you know of a suitable alternative?
Thank you in advance.
All help is much appreciated!
Re: StatFact
HI
i perform Principle component based estimation of static factors from macro-panels, with r determined by Bai and Ng (2002) criteria. My N=39 and T=145.
i would like determine r using my result ( in attachment). How can i do ? (how can i interpret my result)
Many thanks !
, i perform Principle component based estimation of static factors from macro-panels, with r determined by Bai and Ng (2002) criteria. My N=39 and T=145.
i would like determine r using my result ( in attachment). How can i do ? (how can i interpret my result)
Many thanks !
Who is online
Users browsing this forum: No registered users and 31 guests