Near singular matrix

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

dieterE28
Posts: 3
Joined: Fri Mar 20, 2009 3:18 am

Near singular matrix

Postby dieterE28 » Wed May 20, 2009 3:11 am

I am trying to compute the invers of a 900x900 matrix that is of full rank. It contains a lot of zeros, though.
Eviews gives me the "Near singular matrix" error. Is it there an inverse and eviews is just not able to compute it?
If so, is that inverse identical with the Moore-Penrose-Pseudo inverse (which eviews is able to calculate)?

Thanks

trubador
Did you use forum search?
Posts: 1520
Joined: Thu Nov 20, 2008 12:04 pm

Re: Near singular matrix

Postby trubador » Wed May 20, 2009 4:03 am

Moore-Penrose (MP) inverse is a kind of generalized (pseudo) inverse that pre and post multiplying the original matrix by pseudo-inverse returns the inverse. If, however, both columns and rows are linearly independent, then pseudo-inverse will be same as the inverse. EViews can calculate the pseudo-inverse via the following command:

Code: Select all

@pinverse(matrix)


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests