Non-linear least squares regression

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

agamonim
Posts: 4
Joined: Fri Apr 20, 2018 10:58 pm

Non-linear least squares regression

Postby agamonim » Mon Apr 23, 2018 12:11 am

Hi!! I need help on running a non-linear least squares regression. The equation I have is as follows:

Log Y= α+β1 X1*X2+ β2 X3+⋯+ βn Xn+εi
where,
X1=(1-exp⁡(-r×Ti ))*(1/r)

As you can see that one of the independent variables (i.e. X1) is non linear. Besides, it has been interacted with another independent variable (i.e. X2) in this equation. I need to find out the value of parameter "r" here, while the other variables are known to me. Can you please tell me how to run this regression on EViews 6 or on EViews 10 student version?

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13294
Joined: Tue Sep 16, 2008 5:38 pm

Re: Non-linear least squares regression

Postby EViews Gareth » Mon Apr 23, 2018 7:30 am

Follow us on Twitter @IHSEViews

agamonim
Posts: 4
Joined: Fri Apr 20, 2018 10:58 pm

Re: Non-linear least squares regression

Postby agamonim » Sat May 19, 2018 12:12 am

Hi!! I did not find my answer in the link because it is specific.
I am typing the equation in the EViews 6 as follows:

Log Y c F*(1-exp⁡(-r×Ti ))/r X2 X3 X4

The parameter "r" is unknown to me and the non-linear regression result should indicate me what is the value of "r". When I run this regression, I get a notification "R is not defined". How do I define "r"? I know the range of "r" (i.e. between 0.05 to 0.10) but I do not know the exact value that r should take. How do I execute this regression? Can you please help me in this?

startz
Non-normality and collinearity are NOT problems!
Posts: 3775
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear least squares regression

Postby startz » Sat May 19, 2018 7:21 am

Since your equation is nonlinear in parameters you need to write it out in equation form, as shown in the documentation.

The easiest way to indicate that a parameter needs to be estimated is to call it c(1) or c(2), etc. That's how EViews knows it is a parameter rather than a data series.

agamonim
Posts: 4
Joined: Fri Apr 20, 2018 10:58 pm

Re: Non-linear least squares regression

Postby agamonim » Mon May 21, 2018 4:06 am

Thank you!! I have tried it! It works. But again there is a problem of heteroscedasticity because of which I need to use WLS. I want to use inverse of variance of residuals as weights. However, I am little confused as to how to do it. Can you help me in this? I am using version 6 of E Views.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 17 guests