Page 1 of 1

Excluding variables from model solve

Posted: Sun May 12, 2019 6:41 pm
by theagney
Hi there!

I'm trying to exclude an endogenous variable from the model object so that it can then be used as a control variable. When I use the "model.exclude" command, the variable comes up in the model object's exclusion list, but when I go to calculate the control variable (as part of a targeting process), the program throws an error saying that the control variable is still endogenous.

Any thoughts on why it is still endogenous?

Re: Excluding variables from model solve

Posted: Sun May 12, 2019 8:58 pm
by EViews Gareth
Can you provide more precise detail on what you're doing?

Re: Excluding variables from model solve

Posted: Mon May 13, 2019 4:42 pm
by theagney
We have a set of equations used for forecasting economic variables. For some of these equations, we have values we want to target. To target these values, we want to use other equations within this set as controls. To use these equations as controls, we need to exclude them from the model solve (i.e. make them exogenous). When using the "exclude" command, this places the equation that we want as the control into the exclusion list of the scenario, but doesn't comment it out of the text in the model object. Hence when we try to target a variable using the excluded variable as a control (e.g. m.control invest gdp gdp_target), the solver is still saying that the control variable (invest) is still endogenous.

Is the right command/approach for what we are trying to do?

Re: Excluding variables from model solve

Posted: Mon May 13, 2019 4:52 pm
by EViews Gareth
Ah, that is currently not possible the way you're doing it.

You could remove the investment equation, although you'll have to figure out if that breaks other interactions you're interested in.