Effect of RHS variables on dependent variable (_1 vs _0)
Posted: Sat Oct 23, 2010 5:08 pm
Hi,
My model has baseline _0, and scenario _1 [over 2010-2020]. For each equation's dependent variable, I want to see the effect of each right-hand side variable. (For now, assume there are no lagged dependent variables on the rhs).
In other words, suppose I have
LS log(xx) c log(yy(-1)) pdl(zz,6,1,2)
So if xx changes by a certain amount in _1 vs _0, I want to know how much of that came from yy and how much came from zz. I have found no easy way to do this.
So first I needed a list of right-hand variables. I tried eqname.@varlist but that doesn't have all of them. So I converted the equation to a model and got the exogenous variables. Then I set them equal to the _0 values, and looped through them so one at a time they equaled the _1 values, did the equation forecast, and got the result. Is that the only way to do this?
That was for stochastic equations. For an identity in a model, I've found no way to get that text from the model and evaluate it for _0 and _1 values. Any suggestions?
TIA, Dan
My model has baseline _0, and scenario _1 [over 2010-2020]. For each equation's dependent variable, I want to see the effect of each right-hand side variable. (For now, assume there are no lagged dependent variables on the rhs).
In other words, suppose I have
LS log(xx) c log(yy(-1)) pdl(zz,6,1,2)
So if xx changes by a certain amount in _1 vs _0, I want to know how much of that came from yy and how much came from zz. I have found no easy way to do this.
So first I needed a list of right-hand variables. I tried eqname.@varlist but that doesn't have all of them. So I converted the equation to a model and got the exogenous variables. Then I set them equal to the _0 values, and looped through them so one at a time they equaled the _1 values, did the equation forecast, and got the result. Is that the only way to do this?
That was for stochastic equations. For an identity in a model, I've found no way to get that text from the model and evaluate it for _0 and _1 values. Any suggestions?
TIA, Dan