[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
"FEIVESON, ALAN H. (AL) (JSC-SK) (NASA)" <alan.h.feiveson@nasa.gov> |

To |
"'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |

Subject |
st: RE: Could Not Find Idempotent Matrix Inverse |

Date |
Mon, 1 Nov 2004 08:46:49 -0600 |

The only idempotent matrix that is not singular is the identity matrix. All others won't have an inverse. Al Feiveson -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu]On Behalf Of BASSOLE Sent: Sunday, October 31, 2004 12:49 PM To: statalist@hsphsun2.harvard.edu Subject: st: Could Not Find Idempotent Matrix Inverse Dear all, I trying to program to make estimate multi-way error components models with unbalanced data structures. My program is based on Peter DAVIS (2002) article on Journal of Econometric. When I want to make the inverse of Idempotent matrix (Qhhvv), STATA return the error message "Matrix has missing value". But, before making the invers, I check the missings values in the matrix and no missing values are present. After this I decide to calcul the determinant of this matrix, and here STATA return dot, instead of the value of determinant. I dont know why. Before send the mail I make the same program in TSP, and here I have not see the programe for calculate the inverse of the same matrix (Qhhvv). /*Create the orthogonal projection matrices for each variances component*/; *For Household*; mkmat idm1-idm346, matrix(hh); mat Phh = hh*(inv(hh'*hh))*hh'; mat Qhh = INN - Phh; *For Villages*; mkmat codevillage1-codevillage58, matrix(vv); mat Pvv = vv*(inv(vv'*vv))*vv'; mat Qvv = INN - Pvv; /*Apply Lemma 1 (p.70) in Davis(2002)*/; *Here, X1 are households and X2 villages*; mat Qhhvv = Qhh*vv; mat Qhhvv = Qhh*vv; mat PQhhvv = Qhhvv*inv(Qhhvv'*Qhhvv)*Qhhvv'; matrix has missing values r(504); Does anybody know the solution to this problem? Thank you in advance. Léandre BASSOLE CERDI-CNRS 65 Boulevard F. Mitterrand Boite Postale 320 63009 Clermont-Ferrand CEDEX 1 FRANCE Tel : 04 73 17 74 00 Fax : 04 73 17 74 28 * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**st: Heckman with second stage Probit (i.e., Heckprob) and inverse of Mills' Ratio** - Next by Date:
**Re: st: RRR with CI from logit model** - Previous by thread:
**st: Heckman with second stage Probit (i.e., Heckprob) and inverse of Mills' Ratio** - Next by thread:
**st: RE: RE: xscale range with negative upper bound** - Index(es):

© Copyright 1996–2014 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |