@sm_statecov
Posted: Tue Dec 20, 2016 8:00 am
Hi,
I am estimating using the sspace object in Eviews. My question regards the output @sm_statecov. It is my understanding that @sm_statecov is a matrix where each row is a @vech of smoothed state covariance matrix at each point in time (t).
However it seems that I get a negative variance for one of my states in one row of the object.
Then I get a negative variance for state number 2, which can be seen in the element (2,2) of statecov_row2. This only happens for row 2 in the @sm_statecov matrix.
Have I misunderstood what the @sm_statecov output is? And if not, how can the smoothed state variance be negative? The estimation seems to pass through fine.
Thank you in advance!
I am estimating using the sspace object in Eviews. My question regards the output @sm_statecov. It is my understanding that @sm_statecov is a matrix where each row is a @vech of smoothed state covariance matrix at each point in time (t).
However it seems that I get a negative variance for one of my states in one row of the object.
Code: Select all
Matrix statecov = sspace1.@sm_statecov
Vector vector_row2 = @transpose(@Rowextract(statecov,2))
Matrix statecov_row2 = @unvech(vector_row2)
Show statecov_row2
- statecov_row2.jpg (88.24 KiB) Viewed 2785 times
Thank you in advance!