dolfin team mailing list archive

["Garth N. Wells" <gnw20@xxxxxxxxx>]

On 02/06/11 13:41, Anders Logg wrote:
> Anyone using or interested in the ODE solvers should take a look
> below.
> 
> On Thu, Jun 02, 2011 at 02:17:17PM +0200, Benjamin Kehlet wrote:
>> On 2 June 2011 14:02, Anders Logg <logg@xxxxxxxxx> wrote:
>>> On Thu, Jun 02, 2011 at 01:10:01PM +0200, Benjamin Kehlet wrote:
>>>> On 2 June 2011 11:51, Anders Logg <logg@xxxxxxxxx> wrote:
>>>>> On Thu, Jun 02, 2011 at 10:46:29AM +0100, Garth N. Wells wrote:
>>>>>>
>>>>>>
>>>>>> On 02/06/11 10:26, Anders Logg wrote:
>>>>>>> On Thu, Jun 02, 2011 at 10:07:59AM +0100, Garth N. Wells wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>> On 01/06/11 23:46, Anders Logg wrote:
>>>>>>>>> Have you checked that there is no performance penalty?
>>>>>>>>
>>>>>>>> I just have - evaluating a Legendgre polynomial 10k times at the same
>>>>>>>> point is just noise with both methods (of the order 10^-5 - 10^-4 s).
>>>>>>>
>>>>>>> It may be noise for some applications, but not for others. I'm not
>>>>>>> sure this is a bottle-neck for the ODE code (Benjamin will know) but
>>>>>>> we need to evaluate Legendre polynomials of degree > 100 many times
>>>>>>> and then it may not be noise.
>>>>>>>
>>>>>>
>>>>>> For very high degree (e.g. 200) Boost is marginally faster.
>>>>>
>>>>> Sounds promising then.
>>>>>
>>>>>>>> The Boost code is slightly slower because it doesn't cache the values
>>>>>>>> (which is nice not to do), but may be faster if the call is inlined.
>>>>>>>> It's not possible to inline it at the moment because of clashes between
>>>>>>>> tr1:tuple and boost::tuple (Boost bug, I suspect). Old and new are the
>>>>>>>> same when evaluating at different points.
>>>>>>>
>>>>>>> Let's wait for Benjamin to comment.
>>>>>>>
>>>>>>
>>>>>> The speed is about the same (with scope to improve the speed for Boost)
>>>>>> for unique values. The caller should be responsible for caching, if
>>>>>> desired, since it can lead to memory blow out.
>>>>>>
>>>>>> Legendre does not appear in the ode code. It only appears in the
>>>>>> computation of quadrature schemes.
>>>>>
>>>>> True, but the quadrature schemes are used in the ode code.
>>>>>
>>>>>
>>>>>
>>>>>> Garth
>>>>>>
>>>>>>
>>>>>>
>>>>>>>> Garth
>>>>>>>>
>>>>>>>>
>>>>>>>>> Benjamin has
>>>>>>>>> worked quite hard on optimizing some of the basic math routines (in
>>>>>>>>> some cases by many many orders of magnitude).
>>>>>>>>>
>>>>>>>>> Benjamin, can you take a look that it still works?
>>>>
>>>> Yes, the performance seems to be about the same, but I'm unable to
>>>> compile it with support for GMP.
>>>>
>>>> /usr/include/boost/math/special_functions/legendre.hpp:178:
>>>> instantiated from ‘typename boost::math::tools::promote_args<RT,
>>>> float, float, float, float, float>::type boost::math::legendre_p(int,
>>>> int, T, const Policy&) [with T = __gmp_expr<__mpf_struct [1],
>>>> __mpf_struct [1]>, Policy =
>>>> boost::math::policies::policy<boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy,
>>>> boost::math::policies::default_policy>]’
>>>> /usr/include/boost/math/special_functions/legendre.hpp:185:
>>>> instantiated from ‘typename boost::math::tools::promote_args<RT,
>>>> float, float, float, float, float>::type boost::math::legendre_p(int,
>>>> int, T) [with T = __gmp_expr<__mpf_struct [1], __mpf_struct [1]>]’
>>>> /home/benjamik/fenics/dolfin-wells_gmp/dolfin/math/Legendre.cpp:42:
>>>> instantiated from here
>>>> /usr/include/boost/math/special_functions/legendre.hpp:167: error: no
>>>> matching function for call to ‘pow(__gmp_expr<__mpf_struct [1],
>>>> __gmp_binary_expr<long int, __gmp_expr<__mpf_struct [1],
>>>> __gmp_binary_expr<__gmp_expr<__mpf_struct [1], __mpf_struct [1]>,
>>>> __gmp_expr<__mpf_struct [1], __mpf_struct [1]>,
>>>> __gmp_binary_multiplies> >, __gmp_binary_minus> >,
>>>> __gmp_expr<__mpf_struct [1], __gmp_binary_expr<__gmp_expr<__mpf_struct
>>>> [1], __mpf_struct [1]>, long int, __gmp_binary_divides> >)’
>>>> /usr/include/bits/mathcalls.h:154: note: candidates are: double
>>>> pow(double, double)
>>>> /usr/include/c++/4.4/cmath:358: note:                 float
>>>> std::pow(float, float)
>>>> /usr/include/c++/4.4/cmath:362: note:                 long double
>>>> std::pow(long double, long double)
>>>> /usr/include/c++/4.4/cmath:369: note:                 double
>>>> std::pow(double, int)
>>>> /usr/include/c++/4.4/cmath:373: note:                 float std::pow(float, int)
>>>> /usr/include/c++/4.4/cmath:377: note:                 long double
>>>> std::pow(long double, int)
>>>> [...]
>>>>
>>>> boost::math::legendre seems to rely on std::pow which is not
>>>> templated, only implemented with the most common types.
>>>
>>> If it's not possible to make it work, we need to revert back.
>>
>> I don't know of any solution to this. This is the same problem that we
>> discussed some months back (then related to Armadillo): Templated
>> libraries which rely on non-templated  code (often old and implemented
>> i c), so they only support the types which these underlying libraries
>> can handle. I think the only solution here is a change in
>> boost::math::Legendre.
>>
>> Of course another solution would be to split the ODE solver from
>> Dolfin and let it continue as a separate project, and then import code
>> from that when we are going to look at automation/generating code for
>> time-dependent problems.
> 
> Yes, perhaps it's time for that. Since it is going to be removed soon
> (and replaced by code generation), the best option might be to remove
> it before the release of 1.0.
> 
> Are there any objections? Is anyone using the ODE solvers?
> 

No objection, I think that it's a good idea.

Once the ODE solvers are out, we can re-design the arbitrary precision
interface.

Garth

> (They will make a comeback later in new form.)
> 
> --
> Anders