Is it possible to force a coefficient to be a specific value in OLS? For example, if I want C(2) [that is, the coefficient for log(price)] to be -1.5, can I do that?
Here is my equation:
equation {%eqname}
{%eqname}.ls log(ssu) c log(price) log(stores) @trend(1) @expand(period,@dropfirst) easter
Also, is it possible to do this from the PROC menu anywhere or without programming?
Weighting Coefficients
Moderators: EViews Gareth, EViews Moderator
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13600
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Weighting Coefficients
You can always write out an equation as an expression rather than by list (both in a program and via the dialogs).
i.e.
is the same thing as
When written out as an expression you can replace a coefficient with a value, i.e.:
Unfortunately, for you, you cannot use the @expand keyword when writing out an equation by expression, so you'll have to expand it manually.
Of course the alternative, if you want to write out your equation by list, is simply to write your equation as:
i.e.
Code: Select all
ls y c x1 x2Code: Select all
ls y = c(1)+c(2)*x1+c(3)*x2 Code: Select all
ls y = c(1)+0.5*x1 + c(2)*x2Of course the alternative, if you want to write out your equation by list, is simply to write your equation as:
Code: Select all
ls log(ssu)+1.5*log(price) c log(stores) @trend(1) @expand(period,@dropfirst) easter -
cmgconsulting
- Posts: 38
- Joined: Mon Nov 17, 2008 8:25 am
Re: Weighting Coefficients
Can I force a negative sign on C(2).
Is it possible to do this without manually expanding a factor variable?
Is it possible to do this without manually expanding a factor variable?
Who is online
Users browsing this forum: No registered users and 2 guests
