EViews.com

Posted: **Wed Jul 29, 2015 5:14 am**

Good Morning

I want to get all the possible regressions between many independent variables and the same depedent variable.
i am doing the following(i am doing with more betas too, this is just a example):

Code: Select all


for %i x1(-4 to 4) x2(-4 to 4) x3(-4 to 4) x4(-4 to 4) x5(-4 to 4)
xs.add {%i}  
next
group xf
for %f nivel
  xf.add {%f}
 next

 for !f=1 to xf.@count
   %fname = xf.@seriesname(!f)
    for !i=1 to xs.@count-1
       %iname = xs.@seriesname(!i)
          for !j=!i+1 to xs.@count
            %jname = xs.@seriesname(!j)
eq.ls  {%fname} c {%iname} {%jname}
;

but i dont know how to exclude all regressions with multiple equals variables like:
y = x1(-1)+ x1(-3)+x1(-4)
y= x1+x1(4)
....

have someway to do all this regressions without the same variable in the same equation? Anyone can help me with this?

Posted: **Wed Jul 29, 2015 6:44 am**

I'm not sure I follow your question

Posted: **Wed Jul 29, 2015 9:09 am**

lets say that i have GDP, CPI and Interest Rate
and i am trying to explain some variabel y with this 3 variables but i consider diferent lags for them.
so i get:
Y = C + GDP
Y = C + GDP(1)
Y = C + GDP(2)
.
.
.
Y= C + GDP + GDP(1) + GDP(2)
.
.
.
Y= C+ GDP + CPI + CPI(1)
.
.
.
Y= C+ GDP(2) + CPI(-2) + Interest

So i dont want that the same variable with a diferent lag in the same equation like this in red.
i just want combination with diferent variables and your diferent lags

Posted: **Wed Jul 29, 2015 9:12 am**

Code: Select all


create u 100
series gdp=nrnd
series cpi=nrnd
series i=nrnd
series y=nrnd

!n=0
for !i=-4 to 4
	for !j=-4 to 4
		for !k=-4 to 4
			equation eq!n.ls y c gdp(!i) cpi(!j) i(!k)
			!n=!n+1
		next
	next
next

?

Posted: **Wed Jul 29, 2015 9:21 am**

yes, but the problem is that i have 14 variables x and 9 lags for each

Posted: **Wed Jul 29, 2015 9:30 am**

So write a few more for loops.

Posted: **Wed Jul 29, 2015 9:46 am**

Thz man
its works

Have a nice day

Posted: **Wed Jul 29, 2015 10:06 am**

yes, but the problem is that i have 14 variables x and 9 lags for each

Isn't that 14^9 regressions?

Posted: **Wed Jul 29, 2015 10:28 am**

X = independent variable

Posted: **Wed Jul 29, 2015 10:57 am**

9^14?

EViews.com

Multiple regressions with lags

Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags

Re: Multiple regressions with lags