Am 11.11.2024 um 12:32 schrieb mayuda:
Dear all,
We are using the RESET function and when we try to perform Ramsey's
RESET test in Gretl using only the squares version, the result we get
is the one for cubes and squares with robust estimation. We wanted to
know if there's any reason for this or if it's an error.
This looks very much like a bug, which I can confirm here with gretl
2024c on Windows. Note that if you replay the automatically created
command log, where the corresponding command "reset --squares-only"
appears, the wanted test is performed (and yields a different result, of
course). So only when executed from the GUI, gretl is taking a wrong
turn and isn't doing what it's recording in the command log.
Thanks for the report.
Sven