Page 1 of 1
order of RHS variables
Posted: Tue Oct 25, 2011 10:13 am
by startz
It appears that EViews moves AR and MA terms to the end of the equation, no matter what order they're entered in. This is not so good if one is writing a program and wants to control which coefficients go where in the C vector.
Re: order of RHS variables
Posted: Tue Oct 25, 2011 10:25 am
by EViews Gareth
That's the way it has always been. Changing it would be problematic, since there are lots of programs out there that require the ARMA terms to be at the end.
Usually it isn't such a big deal to program around if it does get in your way.
Re: order of RHS variables
Posted: Tue Oct 25, 2011 1:49 pm
by EViews Glenn
It may not be quite as convenient, but if you really wanted control, couldn't you specify by explict expression?
Re: order of RHS variables
Posted: Tue Oct 25, 2011 3:10 pm
by startz
Sure, but it's not so convenient. Not to mention that the program should do what I tell it rather than what it thinks best for me.
Re: order of RHS variables
Posted: Tue Oct 25, 2011 4:04 pm
by EViews Glenn
First off, I agree with you in this case, but Gareth is correct about the problems with changing things at this point since the early TSP code from way back when wasn't designed that way.
But I also want to suggest that your last sentence may be asking for more than you really want
:)
Re: order of RHS variables
Posted: Tue Oct 25, 2011 4:24 pm
by startz
But I also want to suggest that your last sentence may be asking for more than you really want
:)
I have repeatedly asked for a "Do what I would want if I had thought this through more carefully" command. I am still holding out hopes for version 8.