Abstract
The article presents an algorithm for linear regression computations subject to linear parametric equality constraints, linear parametric inequality constraints, or a mixture of the two. No rank conditions are imposed on the regression specification or the constraint specification. The algorithm requires a full Moore-Penrose g-inverse which entails extra computational effort relative to other orthonormalization type algorithms. In exchange, auxiliary statistical information is generated: feasibility of a set of constraints may be checked, estimability of a linear parametric function may be checked, and bias and variance may be decomposed by source.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 59-84 |
| Number of pages | 26 |
| Journal | Journal of Econometrics |
| Volume | 12 |
| Issue number | 1 |
| DOIs | |
| State | Published - Jan 1 1980 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Economics and Econometrics
Cite this
}
Computations for constrained linear models. / Gallant, Andrew Ronald; Gerig, Thomas M.
In: Journal of Econometrics, Vol. 12, No. 1, 01.01.1980, p. 59-84.Research output: Contribution to journal › Article
TY - JOUR
T1 - Computations for constrained linear models
AU - Gallant, Andrew Ronald
AU - Gerig, Thomas M.
PY - 1980/1/1
Y1 - 1980/1/1
N2 - The article presents an algorithm for linear regression computations subject to linear parametric equality constraints, linear parametric inequality constraints, or a mixture of the two. No rank conditions are imposed on the regression specification or the constraint specification. The algorithm requires a full Moore-Penrose g-inverse which entails extra computational effort relative to other orthonormalization type algorithms. In exchange, auxiliary statistical information is generated: feasibility of a set of constraints may be checked, estimability of a linear parametric function may be checked, and bias and variance may be decomposed by source.
AB - The article presents an algorithm for linear regression computations subject to linear parametric equality constraints, linear parametric inequality constraints, or a mixture of the two. No rank conditions are imposed on the regression specification or the constraint specification. The algorithm requires a full Moore-Penrose g-inverse which entails extra computational effort relative to other orthonormalization type algorithms. In exchange, auxiliary statistical information is generated: feasibility of a set of constraints may be checked, estimability of a linear parametric function may be checked, and bias and variance may be decomposed by source.
UR - http://www.scopus.com/inward/record.url?scp=0344907297&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0344907297&partnerID=8YFLogxK
U2 - 10.1016/0304-4076(80)90053-6
DO - 10.1016/0304-4076(80)90053-6
M3 - Article
AN - SCOPUS:0344907297
VL - 12
SP - 59
EP - 84
JO - Journal of Econometrics
JF - Journal of Econometrics
SN - 0304-4076
IS - 1
ER -