Hello,
I want to greate a pseudo out of sample forecast with principal component analysis. Because of this I need a comand to extract the components, after fiting. If I want to get the fit of an normal equation, then I use this way...
equation ar1.ls inflation c inflation(-1)
ar1.fit inflation_hat inflation_se
series inf_hat=inflation_hat
So, fist estimate the principal components and second extract the component one, two, three... as an series.
Is this possible???
Best Regards
Martin
Principal Component
Moderators: EViews Gareth, EViews Moderator, EViews Jason, EViews Matt
-
EViews Glenn
- EViews Developer
- Posts: 2682
- Joined: Wed Oct 15, 2008 9:17 am
Re: Principal Component
To perform this operation interactively, select the Proc/Make Principal Components entry from the group menu (see User's Guide I, p. 403) and fill out the resulting dialog. Simply choose your normalization method and enter a name for each component score series you wish to save (the first component score will be saved into the first name provided, the second into the second name, etc.)
To extract the component scores using a command you should use the makepcomp procedure of a group. For example, if GRP1 has three series X1, X2, and X3, then the command
will place the first component scores computed using the default normalization (normalized loadings, scores have variances equal to the eigenvalue) in the series COMP1. You can use options to change the normalization, etc. Additional details are provided in the Command Reference.
To extract the component scores using a command you should use the makepcomp procedure of a group. For example, if GRP1 has three series X1, X2, and X3, then the command
Code: Select all
grp1.makepcomp comp1
Re: Principal Component
Thank You for answering my requery. I will try it :-)
MartinF
MartinF
Who is online
Users browsing this forum: No registered users and 2 guests
