If I am not mistaken, the function invcdf(X,1,1) reflects a chi-square distribution with 1 degree of freedom and then attempts to compute a value "z" such that

Prob( X <= z) = 1  (because the 3rd argument of the invcdf function is "1").

But then "z" is not a real, number, but infinity.

Alecos Papadopoulos PhD
Affiliate Researcher
Dpt of Economics, Athens University of Economics and Business
Foundation for Economic and Industrial Research (IOBE)
web: alecospapadopoulos.wordpress.com/
ORCID:0000-0003-2441-4550

Hi everybody,

 

also confirmed on Mac Os (intel) with the latest snap.

 

Best regards,

 

Luca

 

Da: Sven Schreiber <sven.schreiber@fu-berlin.de>
Data: lunedì, 4 settembre 2023 10:44
A: gretl-users@gretlml.univpm.it <gretl-users@gretlml.univpm.it>
Oggetto: [Gretl-users] Re: Problem with "invcdf()" and "critical()"

Am 04.09.2023 um 10:33 schrieb f.valentini@pm.univpm.it:
> Dear all,
>
> I would like to report a possible bug in the functions invcdf() and critical().
>
> gretl (2023b, Linux) slows down and crashes (after a while) when I run the commands below.
> <hansl>
> scalar aa = invcdf(X,1,1)
> </hansl>

Confirmed on Windows with 2023b (couldn't test with the latest snapshot
right now) - actually I got a freeze and didn't wait for an eventual crash.

This is potentially a bit of a nasty bug, I guess, since it's a standard
statistical calculation. I wonder when this crept in.

thanks

sven
_______________________________________________
Gretl-users mailing list -- gretl-users@gretlml.univpm.it
To unsubscribe send an email to gretl-users-leave@gretlml.univpm.it
Website: https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/

_______________________________________________
Gretl-users mailing list -- gretl-users@gretlml.univpm.it
To unsubscribe send an email to gretl-users-leave@gretlml.univpm.it
Website: https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/