Hello. Le jeu. 23 juin 2022 à 21:27, Sumanth Rajkumar <rajkumar.suma...@gmail.com> a écrit : > > Hi, > > As part of NUMBERS-187 (Enhanced Support for complex numbers), should we > add support for float types?
I'd say no. At least, any computation should be done with "double" precision. Converting from/to "float" would only be the first/last step during in-place processing of large lists of complex numbers. Hence, it would be oblivious of the "complex" nature of a pair of "double"s, whatever the input/output layout. Anyway, this would seem very low priority (unless you have a use case). > If yes, for operations on float types, can we reuse functions (part of > NUMBERS-188) that operate on double types An example? > and use Java float-double > widening and narrowing conversions? That would be a cast (from "float" to "double" on input and vice-versa on output), no? Regards, Gilles > > https://docs.oracle.com/javase/specs/jls/se10/html/jls-5.html#jls-5.1.2 > > Thanks, > Sumanth --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
