EViews.com

Solved.
Delete it, please.
Regards,

Manuel Palmi

Hi.

Sorry, but I can't understand your question. I don't know what "The "C" or beta overcame the sample and it always until to R751, really I don't know why!" means :(

Hi!

Thank you for your answer. I uploaded a file, really, I don't know how explain it, but for example: If you put in a text

"wf u 1 51"

I'll hope only see 51 "C", betas, constants, intercepts, etc... However I'm seeing 751 (R751) and not 51!

Regards!

When you create a workfile, you define the sample period that does not apply to vectors and matrices. Otherwise, you would not be able to make operations on matrices that have more observations than the sample size. "C" is a predefined (and reserved) coefficient vector of size 751 and gives you the flexibility of running/solving large scale equations/models without worrying about betas. If you ever need a larger coefficent vector, you can always explicitly define your own coefficient vector and use it in your calculations.

Hi,

I don't understand the relation between "my problem" with the matrices. Not only if i write "wf u 1 51" i'll see the same problem, for example if i put File / New / Workfile / Unstructured - Undated -> Observations 51, there is the same problem; with all options (Dated-Regular -> Anual, Monthly, Weekly, etc.), again, there is the same problem, etc, etc.

Anyone know a solution? Maybe I changed a "configuration of default" (I don't remember) but I don't find nothing.

Regards,

Why do you think you have a problem? C is not used for storing data.

Hi again,

For example, in this post where I'm using the "Sample Programs" of Eviews 6. Its in "logl file" -> gprobit1.

============================================================

' GPROBIT1.PRG (1.0 - 6/29/98)
' example program for EViews LogL object
' checked for version 6.0 (3/7/2007)
' grouped (proportions) data probit
'
' Problem 6, page 947 of Greene, William H. (1997) Econometric Analysis,
' 3rd edition, Prentice-Hall

' create workfile
wfcreate gprobit u 10

' fill data
series trucks
trucks.fill 160,250,170,365,210,206,203,305,270,340
series prate
prate.fill 11,74,8,87,62,83,48,84,71,79
prate = prate/100

' set up likelihood for probit
logl ll1
ll1.append @logl logl1
ll1.append xb = c(1)+c(2)*trucks
ll1.append logl1 = prate*log(@cnorm(xb)) + (1-prate)*log(1-@cnorm(xb))

' analytic derivatives
ll1.append @deriv c(1) grad1 c(2) grad2
ll1.append mills1 = @dnorm(xb)/@cnorm(xb)
ll1.append mills2 = -@dnorm(xb)/(1-@cnorm(xb))
ll1.append grad1 = prate*mills1+(1-prate)*mills2
ll1.append grad2 = prate*mills1*trucks+(1-prate)*mills2*trucks

' set OLS starting values
equation eq1.ls prate c trucks

' do MLE
ll1.ml(showopts, m=1000, c=1e-5)
show ll1.output

' create table for answer to problems
table(3,2) ans
setcolwidth(ans,1,30)

' problem 6(a)
setcell(ans,1,1, "answer to 6(a)","l")
scalar ans_a = ( @qnorm(.95)-ll1.c(1) )/ll1.c(2)
ans(1,2) = ans_a

' problem 6(b)
' ans_b is the expected participation rate if budget 6.5 million
setcell(ans,2,1,"expected participation rate","l")
scalar ans_b = @cnorm( ll1.c(1)+ll1.c(2)*650/2 )
ans(2,2) = ans_b

' problem 6(c)
setcell(ans,3,1,"marginal participation rate at 301","l")
scalar ans_c = ll1.c(2)*@dnorm( ll1.c(1)+ll1.c(2)*301 )
ans(3,2) = ans_c

show ans
============================================================

All know which its ok, however, (how you can see the picture) the "Constant" (beta, c, any...) not "take up" the regression, and all seeing:

R1 -2.050451
R2 0.009602
R3 0.000000
R4 0.000000
R5 0.000000
.
.
R751 0.000000

Now, Can you understand me?
Thanks all
Regards,

It is nothing to worry about. C is always exactly the same size in all workfiles.