|
|
Ublas : |
Subject: Re: [ublas] fixed_array in storage.hpp
From: Joaquim Duran Comas (jdurancomas_at_[hidden])
Date: 2015-01-23 18:11:47
Std::array hasn't any constructor defined in C++11, so the initialization
is done as:
std::array<int, 4> v = { { 1, 2, 3, 4} };
C++14 only defines the list constructor to std::array (as you've
wrote: fixed_vector<double,3>
v( {0, 1, 2} ) ;).
Respecting to the "lambda initialiation", I see it as a way to initialize
an array (or a vector, in general) following series like:
i^k: 0, 1, 4, 9, ....
k*i: 0, 2, 4, 6, 8,....
k: 5, 5, 5, 5,....
('k' is a constant value and 'i', the index of the elements of the array)
Joaquim Duran
2015-01-23 2:10 GMT+01:00 Nasos Iliopoulos <nasos_i_at_[hidden]>:
> Thinking it over there may be a reason to implement a fixed_array class
> and that is to remove the default (zero for numeric types) initialization
> of the std::array. This would speed up code like:
>
> (using ublas::fixed_vector here:)
>
> fixed_vector<double,3 > v; // Right now this will initialize to 3 zeros
> which is redundant.
>
> for ( auto i = 0 ; i != 3; i++)
> {
> v[i] = i;
> }
>
> For small arrays though this there is a safe and much better alternative:
>
> fixed_vector<double,3> v( {0, 1, 2} ) ;
>
> and hence I am not sure if it would be wise to implement a
> non-initialization fixed_array class.
>
> Another option I was thinking was to implement some sort of "lambda
> initialization" i.e. a user provided function that is used in the
> constructor. The problem with this is that it may provide opportunities to
> break encapsulation and that would be terrible in the construction phase. I
> am not ruling out such a feature though.
>
> -Nasos
>
> On 01/22/2015 05:59 PM, Joaquim Duran Comas wrote:
>
> I see. I can use std:array.
>
>
> 2015-01-22 15:03 GMT+01:00 Nasos Iliopoulos <nasos_i_at_[hidden]>:
>
>> Why do you want to do that? Can't you use std::array?
>>
>> On 01/21/2015 08:04 PM, Joaquim Duran Comas wrote:
>>
>> I would like to create a fixed_array in strorage.hpp file. Is there any
>> inconvenient?
>>
>> Joaquim Duran
>>
>>
>>
>> _______________________________________________
>> ublas mailing listublas_at_[hidden]http://lists.boost.org/mailman/listinfo.cgi/ublas
>> Sent to: athanasios.iliopoulos.ctr.gr_at_[hidden]
>>
>>
>>
>> _______________________________________________
>> ublas mailing list
>> ublas_at_[hidden]
>> http://lists.boost.org/mailman/listinfo.cgi/ublas
>> Sent to: jdurancomas_at_[hidden]
>>
>
>
>
> _______________________________________________
> ublas mailing listublas_at_[hidden]http://lists.boost.org/mailman/listinfo.cgi/ublas
>
> Sent to: nasos_i_at_[hidden]
>
>
>
> _______________________________________________
> ublas mailing list
> ublas_at_[hidden]
> http://lists.boost.org/mailman/listinfo.cgi/ublas
> Sent to: jdurancomas_at_[hidden]
>