ordered model

[maxchen]

Hi, there may be some bugs here (ver 2010-11-11)

Code: Select all

%ex7 = "\Example Files\EV7 Manual Data"
%ch26 = "\Chapter 26 - Discrete and Limited Dependent Variable Models"
%wf = @evpath + %ex7 +%ch26 +"\order.wf1"
wfopen %wf

equation eq01.ordered danger body brain sleep
c = 0
equation eq02.ordered(showopts) danger body brain sleep
c = 0
equation eq03.ordered(showopts) danger = c(1)*body + c(2)*brain + c(3)*sleep   'by expression
freeze(gfr)   eq02.resids   


then,
1 eq03 usese the current values in C as starting values, furthermore, the output does not report the limit
points
2 in graph gfr, the fitted value seems to be residuals+actual value, it seems no meaning

what's more, by point and click, select eq02, View -> Generalize residual graph, then the legend of the graph reads "standardized residual" not "generalized residual" (it is fine with V5.1)
one last concern, Eq26.23 on user manual, the gamma_{y_{i+1}} should be gamma_{y_{i}+1}. Note that y_{i+1} is not equal to y_{i}+1

[maxchen]

hmm, just now I've found that the following command will cause error

Code: Select all

eq02.makelimits(v) lv   'cov of limit points


if I remove the comment, it will work

[EViews Glenn]

1. The eq03 *should* use the workfile starting values. All estimation using expression in EViews uses the workfile C...As to the display of coefficients, it's a bug that we introduced whilst fixing another one. We'll get a fix posted for it.

2. You are correct, the command should be disallowed for this view. Note that it is correctly removed from the interactive menus.

3. We'll take a look at that one. I'm surprised that it's changed from v6.

4. We'll fix the notation. Thanks for pointing this out.