R-squared in my results

[Zoeltje]

Dear all,

I'm using the following in an Eviews Program:

matrix(3460, 4) results

for !i = 1 to 3460

c(1)=0.1
c(2)=0.00001
c(3)=0.1
c(4)=1.9

equation test.LS(n) RST_DIVC!i*XT_DIVC!i = c(1)+ ((1+EXP(-@abs(C(2))*((c(3)-c(4))*XT_DIVC!i(-3)*(XT_DIVC!i(-1)-(RST_DIVC!i*XT_DIVC!i(-2))))))^-1)*C(3)*XT_DIVC!i(-1)+(1-((1+EXP(-@abs(C(2))*((c(3)-c(4))*XT_DIVC!i(-3)*(XT_DIVC!i(-1)-(RST_DIVC!i*XT_DIVC!i(-2))))))^-1))*C(4)*XT_DIVC!i(-1)

for !j=1 to 4
results(!i, !j) = c(!j)

next
next


But now I not only want to get the coefficients in my results, but also the R-sqaured of each equation (I have 3460 different equations). Does anybody know what I should do? I think I have to use the "@r2", but I don't know how. I tried some things, but they all didn't work.

Thanks in advance!

[EViews Glenn]

I just modified the results matrix to have 5 instead of 4 columns, and added a line at the bottom of the loop to put the r2 inthe 5th column. Note also that I've changed the coefficient references inside the loop to be explicit about their source (equation TEST):

Code: Select all

matrix(3460, 5) results for !i = 1 to 3460 c(1)=0.1 c(2)=0.00001 c(3)=0.1 c(4)=1.9 equation test.LS(n) RST_DIVC!i*XT_DIVC!i = c(1)+ ((1+EXP(-@abs(C(2))*((c(3)-c(4))*XT_DIVC!i(-3)*(XT_DIVC!i(-1)-(RST_DIVC!i*XT_DIVC!i(-2))))))^-1)*C(3)*XT_DIVC!i(-1)+(1-((1+EXP(-@abs(C(2))*((c(3)-c(4))*XT_DIVC!i(-3)*(XT_DIVC!i(-1)-(RST_DIVC!i*XT_DIVC!i(-2))))))^-1))*C(4)*XT_DIVC!i(-1) for !j=1 to 4 results(!i, !j) = test.c(!j) next results(!i, 5) = test.@r2 next

[Zoeltje]

Thanks! That works. I only get the message "Equation estimates are invalid or nonexistent in "RESULTS(10, 5) = TEST.@R2" in a 141 times. For those 141 equations the R-squared is zero. Does somebody know what's wrong with those equations?

Thanks in advance.

[EViews Glenn]

If you save an equation each time through the loop, you can look at that equation in isolation after the fact.