Does anybody happen to have a working rational number library in D? Any plans to add that to Phobos at some point?

Just wondering, because I'm working on some numerical code that would benefit from a custom number type that could be implemented in terms of rational numbers. I just don't feel like reinventing a rational library if one already exists. :)


T

-- 
GEEK = Gatherer of Extremely Enlightening Knowledge

On Wednesday, 25 September 2013 at 16:59:15 UTC, H. S. Teoh wrote:
> Does anybody happen to have a working rational number library in D?

https://github.com/p0nce/gfm/blob/master/gfm/math/fraction.d

disclaimer: I've never used it for anything except writing its unittest block.

On Wednesday, 25 September 2013 at 17:15:25 UTC, ponce wrote:
> On Wednesday, 25 September 2013 at 16:59:15 UTC, H. S. Teoh wrote:
>> Does anybody happen to have a working rational number library in D?

Also David Simcha wrote a better one: https://github.com/dsimcha/Rational/blob/master/rational.d

On 25/09/13 19:21, ponce wrote:
> On Wednesday, 25 September 2013 at 17:15:25 UTC, ponce wrote:
>> On Wednesday, 25 September 2013 at 16:59:15 UTC, H. S. Teoh wrote:
>>> Does anybody happen to have a working rational number library in D?
>
> Also David Simcha wrote a better one:
> https://github.com/dsimcha/Rational/blob/master/rational.d

Why was this never incorporated into Phobos?

Joseph Rushton Wakeling:

> Why was this never incorporated into Phobos?

First it needs the normal review process for Phobos modules :) It should go in the review queue, if it's good enough for Phobos.

> Rational!(BigInt) getTerm(int termNumber) {

I suggest to add inside that "rational.d" module an alias with Bigint, as:

alias Fraction = Rational!BigInt;

Because I don't think I want to use a rational type made with built-in integrals.

Bye,
bearophile

On 26/09/13 16:09, bearophile wrote:
> First it needs the normal review process for Phobos modules :) It should go in
> the review queue, if it's good enough for Phobos.

Fair enough, I was wondering if there was any reason David never submitted it.

> I suggest to add inside that "rational.d" module an alias with Bigint, as:
>
> alias Fraction = Rational!BigInt;
>
> Because I don't think I want to use a rational type made with built-in integrals.

Good call.

On 9/26/13 7:00 AM, Joseph Rushton Wakeling wrote:
> On 25/09/13 19:21, ponce wrote:
>> On Wednesday, 25 September 2013 at 17:15:25 UTC, ponce wrote:
>>> On Wednesday, 25 September 2013 at 16:59:15 UTC, H. S. Teoh wrote:
>>>> Does anybody happen to have a working rational number library in D?
>>
>> Also David Simcha wrote a better one:
>> https://github.com/dsimcha/Rational/blob/master/rational.d
>
> Why was this never incorporated into Phobos?

Probably because it wasn't proposed for review. I'm favorable to integrating it.

Andrei

On Thu, Sep 26, 2013 at 04:50:47PM -0700, Andrei Alexandrescu wrote:
> On 9/26/13 7:00 AM, Joseph Rushton Wakeling wrote:
> >On 25/09/13 19:21, ponce wrote:
> >>On Wednesday, 25 September 2013 at 17:15:25 UTC, ponce wrote:
> >>>On Wednesday, 25 September 2013 at 16:59:15 UTC, H. S. Teoh wrote:
> >>>>Does anybody happen to have a working rational number library in D?
> >>
> >>Also David Simcha wrote a better one: https://github.com/dsimcha/Rational/blob/master/rational.d
> >
> >Why was this never incorporated into Phobos?
> 
> Probably because it wasn't proposed for review. I'm favorable to integrating it.
[...]

Me too.


T

-- 
2+2=4. 2*2=4. 2^2=4. Therefore, +, *, and ^ are the same operation.

On 9/26/2013 12:49 PM, Joseph Rushton Wakeling wrote:
> On 26/09/13 16:09, bearophile wrote:
>> First it needs the normal review process for Phobos modules :) It should go in
>> the review queue, if it's good enough for Phobos.
>
> Fair enough, I was wondering if there was any reason David never submitted it.

For things to happen, usually a self-appointed champion is needed that will relentless push it forward. Things tend to fall by the wayside otherwise.

On 27/09/13 09:36, Walter Bright wrote:
> For things to happen, usually a self-appointed champion is needed that will
> relentless push it forward. Things tend to fall by the wayside otherwise.

In the past David has seemed like a pretty reliable champion of his own work -- quite a few of the parts of Phobos I use regularly are down to him!

I've written to him asking for his input on std.rational but if he doesn't have time/inclination to pursue it, I'll take a look at it.

At first glance what's there looks pretty good with the main issues being stylistic rather than content-related.