Several Tests
Posted: Fri Aug 21, 2015 2:55 am
Hello Guys, im trying to chose the model that best fit to my data, and im doing all possible regression and keeping all the vallues that mean to me, but miss some tests and im want to know how i can do this in the same program:
[*] scalled coefficients[*]
[*] variance inflation factor[*]
[*]correlogram q-stats[*]
[*]white heteroskedasticity[*]
its possible to get vallues from this tests?
in this momment im doing this (exemple for only 1 beta, but im also doing for more betas)
[*] scalled coefficients[*]
[*] variance inflation factor[*]
[*]correlogram q-stats[*]
[*]white heteroskedasticity[*]
its possible to get vallues from this tests?
in this momment im doing this (exemple for only 1 beta, but im also doing for more betas)
Code: Select all
'create empty equation to be used inside the loop
equation eq
'create vector to store regressors name
svector(115) specs1
matrix(115,7) grupo1
!rowcounter=1
for !f=1 to xf.@count
%fname = xf.@seriesname(!f)
for !i=1 to xs.@count
%iname = xs.@seriesname(!i)
eq.ls {%fname} c {%iname}
specs1(!rowcounter) = eq.@spec
!rowcounter = !rowcounter+1
freeze(eqtable) eq
grupo1(!rowcounter-1,1) = @val(eqtable(9,2))
grupo1(!rowcounter-1,2) = @val(eqtable(10,2))
grupo1(!rowcounter-1,3) = @val(eqtable(9,5))
grupo1(!rowcounter-1,4) = @val(eqtable(10,5))
grupo1(!rowcounter-1,5) = @val(eqtable(12,2))
grupo1(!rowcounter-1,6) = @val(eqtable(13,2))
grupo1(!rowcounter-1,7) = @val(eqtable(17,5))
delete eqtable
next
next