Hi Luis, ci() was introduced as part of the native-complex changes (obviously). It should be considered temporary for now, since the current way forward I see is to integrate PDL::Complex into the native stuff (since they actually appear to be binary-compatible, as mentioned earlier), and therefore be able to use the “i” constant in the same way as currently. However, are you seeing a use for it that adds anything to simply using i2C from PDL::Complex?
Best regards, Ed From: Luis Mochan<mailto:moc...@icf.unam.mx> Sent: 07 March 2021 22:43 To: pdl-devel@lists.sourceforge.net<mailto:pdl-devel@lists.sourceforge.net>; perldl<mailto:pdl-gene...@lists.sourceforge.net> Subject: Re: [Pdl-devel] PDL 2.027 released Hello, I'm intrigued by ci. What is the reason for its behavior? I found that ci was defined through pp_def with one output argument. Thus, ci+$x is interpreted as ci(+$x) which actually assigns complex i to each element of $x if $x is a complex pdl. If $x is real, then it assigns 0 (I guess it type-converts i though its real part). Is there a case where you would expect a call to ci with an actual argument? Would it be reasonable to flag an error if it is called with a non-null argument? Would this be costly? My guess is that ci+$x or ci*$x would become a frequenct headache with the actual behavior. Or maybe one can keep the current function ci and include an actual constant use constant CI=>ci; and encourage the use of the constant CI unless there is a good reason to use ci directly. The conversion back and forth between pairs of real numbers and complex numbers without requiring actual multiplication with ci would also be convenient. Given that C99 complex numbers are an array of two reals, it might be simple. Regards, Luis -- o W. Luis Mochán, | tel:(52)(777)329-1734 /<(*) Instituto de Ciencias Físicas, UNAM | fax:(52)(777)317-5388 `>/ /\ Av. Universidad s/n CP 62210 | (*)/\/ \ Cuernavaca, Morelos, México | moc...@fis.unam.mx /\_/\__/ GPG: 791EB9EB, C949 3F81 6D9B 1191 9A16 C2DF 5F0A C52B 791E B9EB _______________________________________________ pdl-devel mailing list pdl-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pdl-devel
_______________________________________________ pdl-devel mailing list pdl-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pdl-devel
