Hi Guillermo, This is unfortunately an example of why one should (almost) never do exact equality checks with floating-point numbers, which are unfortunately rarely precisely the quantity we think of them as. PDL’s “power” operation uses the “pow” libm function (or rather, since we use “tgmath” to simplify the code, really the “pow” family of functions), and is entirely subject to its whims:
pdl> p $x=pdl(-3) -3 pdl> p $x**2-9 0 pdl> p $z=-3+0*i -3 pdl> p $z->info PDL: CDouble D [] pdl> p $x->info PDL: Double D [] pdl> p $z**2-9 1.77635683940025e-15-2.20436423846524e-15i The “info” calls show the types, which are as expected. If you want a complex -3, you can just do cdouble(-3). I infer that “pow”, with complex numbers, may convert to a polar format, and a tiny instability is introduced. But the real solution here is to not use ==, but instead to do something like: If (abs($quantity1-$quantity2) < $tolerance) ... Best regards, Ed From: Guillermo P. Ortiz<mailto:gor...@exa.unne.edu.ar> Sent: 13 April 2023 15:05 To: pdl-de...@lists.sourceforge.net<mailto:pdl-de...@lists.sourceforge.net>; pdl-general@lists.sourceforge.net<mailto:pdl-general@lists.sourceforge.net> Subject: [Pdl-devel] unexpected results Dear pdl folks, doing $x=pdl(-3) with PDL v2.081 we obtain $x**2-9==0 but, defining $z=-3+0*i $z**2-9 =! 0 while $z*$z-9 ==0 May be integer power are computed differently for real but not for intrinsic complex? (logarithmically for any power) Regards -- Dr. Guillermo P. Ortiz Electromagnetismo Aplicado Dto. Física, Facultad de Ciencias Exactas Universidad Nacional del Nordeste Avda Libertad 5460, W3404AAS Corrientes, Argentina. (+54) 379-4424678 interno 4613 gortiz at unne edu ar
_______________________________________________ pdl-general mailing list pdl-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pdl-general
