Re: [racket-dev] Racket Questions?
I'll implement `flmodulo' in the math library, which already has `flodd?' and `fleven?'. These things can be tricky to get right; for example, most people wouldn't notice that their implementation is wrong on negative inputs. I'll probably also do `flremainder'. Neil ⊥ On 09/15/2012 12:45 PM, Michael Wilber wrote: For the record, I've always just defined my own modulo when I need it for floats: ; A modulo operator for floats! (define (float-modulo p q) (- p (* q (truncate (/ p q) It doesn't properly handle negative numbers though. David Van Horn dvanh...@ccs.neu.edu writes: On 9/14/12 3:36 PM, Becca MacKenzie wrote: Hello! So a friend of mine just started learning Racket and was wondering if there's a particular reason why the modulo function in racket only takes in integers? He wrote his own mod function to take in other things but he was just wondering what the reasoning is behind this. Hi Becca, Excellent question -- I hope you don't mind that I've forwarded it to the Racket developers list for a more authoritative answer (and potentially a change to Racket). I don't believe there's any principled reason not to extend `modulo' to other kinds of numbers such as rationals and (exact) complex numbers. I worry that the idea of modulo may not be well defined for inexact numbers, but I could be wrong (inexact numbers don't obey a lot of the usual mathematical properties we're used to). I see that in Mathematica, the arguments of Mod can be any numeric quantities, not necessarily integers. Here are some examples: http://reference.wolfram.com/mathematica/ref/Mod.html#6881 Recently, Racket's GCD and LCM were extended to work on non-integer arguments, and I believe this is a similar case where the function could (and should?) be extended to work for more kinds of numbers. But I'm interested to hear what the dev list has to say on the matter. David _ Racket Developers list: http://lists.racket-lang.org/dev _ Racket Developers list: http://lists.racket-lang.org/dev _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] Racket Questions?
On 9/14/12 3:36 PM, Becca MacKenzie wrote: Hello! So a friend of mine just started learning Racket and was wondering if there's a particular reason why the modulo function in racket only takes in integers? He wrote his own mod function to take in other things but he was just wondering what the reasoning is behind this. Hi Becca, Excellent question -- I hope you don't mind that I've forwarded it to the Racket developers list for a more authoritative answer (and potentially a change to Racket). I don't believe there's any principled reason not to extend `modulo' to other kinds of numbers such as rationals and (exact) complex numbers. I worry that the idea of modulo may not be well defined for inexact numbers, but I could be wrong (inexact numbers don't obey a lot of the usual mathematical properties we're used to). I see that in Mathematica, the arguments of Mod can be any numeric quantities, not necessarily integers. Here are some examples: http://reference.wolfram.com/mathematica/ref/Mod.html#6881 Recently, Racket's GCD and LCM were extended to work on non-integer arguments, and I believe this is a similar case where the function could (and should?) be extended to work for more kinds of numbers. But I'm interested to hear what the dev list has to say on the matter. David _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] Racket Questions?
For the record, I've always just defined my own modulo when I need it for floats: ; A modulo operator for floats! (define (float-modulo p q) (- p (* q (truncate (/ p q) It doesn't properly handle negative numbers though. David Van Horn dvanh...@ccs.neu.edu writes: On 9/14/12 3:36 PM, Becca MacKenzie wrote: Hello! So a friend of mine just started learning Racket and was wondering if there's a particular reason why the modulo function in racket only takes in integers? He wrote his own mod function to take in other things but he was just wondering what the reasoning is behind this. Hi Becca, Excellent question -- I hope you don't mind that I've forwarded it to the Racket developers list for a more authoritative answer (and potentially a change to Racket). I don't believe there's any principled reason not to extend `modulo' to other kinds of numbers such as rationals and (exact) complex numbers. I worry that the idea of modulo may not be well defined for inexact numbers, but I could be wrong (inexact numbers don't obey a lot of the usual mathematical properties we're used to). I see that in Mathematica, the arguments of Mod can be any numeric quantities, not necessarily integers. Here are some examples: http://reference.wolfram.com/mathematica/ref/Mod.html#6881 Recently, Racket's GCD and LCM were extended to work on non-integer arguments, and I believe this is a similar case where the function could (and should?) be extended to work for more kinds of numbers. But I'm interested to hear what the dev list has to say on the matter. David _ Racket Developers list: http://lists.racket-lang.org/dev _ Racket Developers list: http://lists.racket-lang.org/dev
