Re: [Quantlib-dev] math constants

[Peter C. <pca...@gm...>]

this is what is shipped with gcc

/* Some useful constants.  */
#if defined __USE_MISC || defined __USE_XOPEN
# define M_E 2.7182818284590452354 /* e */
# define M_LOG2E 1.4426950408889634074 /* log_2 e */
# define M_LOG10E 0.43429448190325182765 /* log_10 e */
# define M_LN2 0.69314718055994530942 /* log_e 2 */
# define M_LN10 2.30258509299404568402 /* log_e 10 */
# define M_PI 3.14159265358979323846 /* pi */
# define M_PI_2 1.57079632679489661923 /* pi/2 */
# define M_PI_4 0.78539816339744830962 /* pi/4 */
# define M_1_PI 0.31830988618379067154 /* 1/pi */
# define M_2_PI 0.63661977236758134308 /* 2/pi */
# define M_2_SQRTPI 1.12837916709551257390 /* 2/sqrt(pi) */
# define M_SQRT2 1.41421356237309504880 /* sqrt(2) */
# define M_SQRT1_2 0.70710678118654752440 /* 1/sqrt(2) */
#endif

/* The above constants are not adequate for computation using `long double's.
   Therefore we provide as an extension constants with similar names as a
   GNU extension.  Provide enough digits for the 128-bit IEEE quad.  */
#ifdef __USE_GNU
# define M_El 2.718281828459045235360287471352662498L /* e */
# define M_LOG2El 1.442695040888963407359924681001892137L /* log_2 e */
# define M_LOG10El 0.434294481903251827651128918916605082L /* log_10 e */
# define M_LN2l 0.693147180559945309417232121458176568L /* log_e 2 */
# define M_LN10l 2.302585092994045684017991454684364208L /* log_e 10 */
# define M_PIl 3.141592653589793238462643383279502884L /* pi */
# define M_PI_2l 1.570796326794896619231321691639751442L /* pi/2 */
# define M_PI_4l 0.785398163397448309615660845819875721L /* pi/4 */
# define M_1_PIl 0.318309886183790671537767526745028724L /* 1/pi */
# define M_2_PIl 0.636619772367581343075535053490057448L /* 2/pi */
# define M_2_SQRTPIl 1.128379167095512573896158903121545172L /* 2/sqrt(pi) */
# define M_SQRT2l 1.414213562373095048801688724209698079L /* sqrt(2) */
# define M_SQRT1_2l 0.707106781186547524400844362104849039L /* 1/sqrt(2) */
#endif

On 13 November 2015 at 17:31, Luigi Ballabio <lui...@gm...> wrote:
> Yes, it might make sense.  But truth be told, I'd be for having less of them
> rather than more.  Which of the macros are already defined by the compilers
> we support?
>
> Luigi
>
>
> On Wed, Nov 4, 2015 at 10:38 AM Peter Caspers <pca...@gm...>
> wrote:
>>
>> Hello Luigi, all,
>>
>> I had a typo in my code and used M_SQRT_2 instead of M_SQRT2, the
>> former being defined as 1 / sqrt(2), but what I wanted was sqrt(2). We
>> also have a macro M_SQRT1_2 with the same value and which seems to
>> have a better name. I find M_SQRT_2 quite confusing and non-intuitive,
>> so I am wondering if we could just deprecate it and remove it in one
>> of the next versions ?
>>
>> There are also some constants whose meaning does not appear
>> crystal-clear at first sight (M_IVLN10, M_LN2LO, M_LN2HI, M_INVLN2).
>> Can someone provide a definition, so that we can maybe add them to the
>> source ?
>>
>> Finally there are some constants with many digits (M_SQRT1_2, M_PI).
>> Since there is no specifier (L) I guess they are always mapped to
>> double by the compiler, which should be 64bit on every platform, so
>> effectively they won't help even if one uses them to initialize a
>> __float128 etc. ? Maybe we should rather do it like GNU and add
>> versions of the macros like M_PIl, M_SQRT1_2l defining more digits and
>> with the L suffix ? Then we would also be in line with gcc.
>>
>> What do you think ?
>>
>> Best regards
>> Peter
>>
>>
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> QuantLib-dev mailing list
>> Qua...@li...
>> https://lists.sourceforge.net/lists/listinfo/quantlib-dev
>
> --
>
> <http://leanpub.com/implementingquantlib>
> <http://implementingquantlib.com>
> <http://twitter.com/lballabio>