[Gretl-devel] bug which is new in 1.7.4

[Riccardo (Jack) Lucchetti]

On Thu, 3 Apr 2008, Allin Cottrell wrote:

> The simple fix, I think, is not to decrement i until we know that
> the variable in question really has been deleted.  However, this
> issue raises the question of whether the last model estimated
> within a function that is returning should really be treated as
> the "last model".
>
> For now, I'm committing the "simple fix" to CVS just so you can
> take a look.

I can confirm that the "simple fix" lets Sven's script and my example 
script run, but the problem remains of the workspace being contaminated by 
several local variables being promoted to global.

I'm in favour of removing from the stack all locally estimated models. The 
only thing we lose is not being able to use dollar accessors immediately 
after a function, eg in

<script>
function foo(series x, series y)
     var 1 x y
end function

nulldata 10
x = normal()
y = normal()
foo(x,y)
A = $coeff
print A
</script>

which works now, but wouldn't after my proposed change. But I'm personally 
ok with this, it's bad syntax anyway.


Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche

r.lucchetti at univpm.it
http://www.econ.univpm.it/lucchetti