Re: [Gretl-users] bug in string substitution?

I spoke to soon in my earlier message -- the problem now seems even stranger. The attached script embeds the string substitution part into the loops that generate the vector of dates. Now it seems that the substitution works mostly fine -- the following sequence gets generated in the loops: 1986.05 1986.06 1986.07 1986.08 1986.09 1986.1 1986.11 1986.12 1987.01 1987.02 1987.03 So the "10" is still not behaving properly. But the vector that gets built comes out like this: 1986.0 1986.0 1986.1 1986.1 1986.1 1986.1 1986.1 1986.1 1986.1 1987.0 1987.0 1987.0 Very strange... PS -----Original Message----- From: gretl-users-bounces(a)lists.wfu.edu [mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell Sent: Friday, June 14, 2013 9:39 AM To: Gretl list Subject: Re: [Gretl-users] bug in string substitution? On Thu, 13 Jun 2013, Summers, Peter wrote: Hmm, I'm not seeing a problem here (current CVS on Linux). But note that the recommended way to get a month to appear in the desired format with sprintf is to use the conversion "%02d" (which inserts a zero if and only if it's needed). If I do <hansl> scalar yr = 2009 scalar mo = 4 string s sprintf s, "%d.%02d", yr, mo printf "s = '%s'\n", s scalar d = @s </hansl> I get: <output> ? scalar yr = 2009 Generated scalar yr = 2009 ? scalar mo = 4 Generated scalar mo = 4 ? string s ? sprintf s, "%d.%02d", yr, mo s = '2009.04' ? scalar d = 2009.04 Generated scalar d = 2009.04 </output> Allin Cottrell _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users