(no subject) - Gretl-users - gretlml.univpm.it

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

(no subject)

MLE with endogenouly determined...

bugs in gretl 1.9.5cvs

cox m. (mc12v07)

Wednesday, 20 July 2011 Wed, 20 Jul '11

8:31 a.m.

Hi, I am trying to perform a constrained linear regression in gretl for which the beta coefficients are all non-negative and sum to unity. Is there anyway that I can do this? Kind Regards Matt

Reply

Show replies by thread

Sven Schreiber

Wednesday, 20 July Wed, 20 Jul

7:49 p.m.

Am 20.07.2011 10:31, schrieb cox m. (mc12v07):

Hi, I am trying to perform a constrained linear regression in gretl for which the beta coefficients are all non-negative and sum to unity. Is there anyway that I can do this?

I would suggest a simple approach first, i.e. express the third coefficient as 1-b1-b2, use NLS, and only verify afterwards whether all coefficients are non-negative. If you get some negative coefficients, these might be set to zero, or you can then think about more sophisticated solutions :-) cheers, sven

Reply

4663

days inactive

4663

days old

gretl-users@gretlml.univpm.it

Manage subscription

1 comments

2 participants

tags (0)

participants (2)

cox m. (mc12v07)
Sven Schreiber