On Mon, Oct 13, 2025 at 9:09 AM Roland Clarke <rolandclarke97(a)icloud.com> wrote:
Hi Allin,
I realise that I didn’t give all the information. I attach a zipped minimal version, in
which the source line has been replaced by the explicit R function which I use to retrieve
the data. You will see that the initial dataset contains three string series in which the
first variable SLIST is a list of variables to be downloaded from the FRED database. I
have reduced this to three, and taken out other irrelevant material in the script. The
datafile from which the script reads the variables is in the directory test/data.
As I mentioned this script works perfectly on a Mac with Gretl 2025b but fails in Ubuntu
24.04 with Gretl 2023c.
I hope this gives the necessary information.
It does indeed, thanks!
I find that your test script runs fine, both with current gretl and
with gretl 2023c as built today from the historic sources using
current gcc. I'm afraid I'm not in a position to test with the
original Debian build of 2023c, but actually there wouldn't be much
point in that since there's nothing we can do to fix a release from
two years ago, even if we could replicate a problem and figure out its
cause.
Anyway, I hope that Jack's new hansl code does what you need.
Allin