- Boost - lists.boost.org

[GSoC][Trie][Radix sort]
by matijas9 19 Mar '14

19 Mar '14

Hello, On list of items for Google Summer of Code is "Trie data structure, extending the work done during GSoC 2013". I would like to know what exactly mean "extending the work", i.e. what is to do in implementation of trie from previous GSoC. Secondly, I would like to add to my proposition of project for GSoC radix sort algorithm. I would like to know how generalized can be types of input datas for radix sort. Elements of sequences which are sorted must belong to some finite universe and should be comparable in constant time, so in my opinion radix sort can sort only sequences of natural numbers well. Therefore I think about preparing version of radix sort sorting sequences of numbers and another version sorting strings (which can be interpreted as sequences of numbers). Can someone think of some other, usefull types of objects that can be sorted using radix sort? Best Regards, Mateusz Wasylkiewicz

2 1

[Parallel Algorithms] Queries regarding the Boost.Thread / Parallel algorithms
by Naveen 18 Mar '14

18 Mar '14

Hi, I am interested in the implementation of parallel algorithms. From given documentation in http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2012/n3429.pdf I could understand the basic crux of the project. But the part which I was not able to figure from the documentation are these; 1. The type of network layer or the communication layer to be used. Can we make use of GASNet or ARMCI as the communication layer and build of ideas on it? Or do we have any specific library to work on this? 2. And the implementation of parallel_for and parallel_for_each, are we trying to mimic the available functionality from the parallel programming models like OpenMP? How far are these functions different from them. 3. Should we analyze about compiler optimizations. Because once these external parallel libraries are called; this necessarily means all the compiler optimizations within this parallel region should be excluded or else the complexity of the parallel library would be increased. Please confirm whether, my understanding is correct or am i looking at the project in a completely unrelated way. Regards, *NAVEEN* | Mobile: 832-720-2393 | about.me <http://about.me/naveen.namashivayam> |

1 0

Queries regarding the Boost.Thread / Parallel algorithms
by Naveen 18 Mar '14

18 Mar '14

Hi, I am interested in the implementation of parallel algorithms. From given documentation in http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2012/n3429.pdf I could understand the basic crux of the project. But the part which I was not able to figure from the documentation are these; 1. The type of network layer or the communication layer to be used. Can we make use of GASNet or ARMCI as the communication layer and build of ideas on it? Or do we have any specific library to work on this? 2. And the implementation of parallel_for and parallel_for_each, are we trying to mimic the available functionality from the parallel programming models like OpenMP? How far are these functions different from them. 3. Should we analyze about compiler optimizations. Because once these external parallel libraries are called; this necessarily means all the compiler optimizations within this parallel region should be excluded or else the complexity of the parallel library would be increased. Please confirm whether, my understanding is correct or am i looking at the project in a completely unrelated way. Regards, *NAVEEN* | Mobile: 832-720-2393 | about.me <http://about.me/naveen.namashivayam> |

1 0

Re: [boost] Welcome as organization member - HELP !!!!!
by Vicente J. Botet Escriba 18 Mar '14

18 Mar '14

Hi, I have reached to see the projects now, Thanks, Vicente Le 18/03/14 22:27, Niall Douglas a écrit : > Ability to assign mentors opened today, I've assigned all the mentors > to projects in so far as I know which. > > Niall > > > On 18 Mar 2014 at 20:56, Boris Schäling wrote: > >> >>> -----Original Message----- >>> From: Niall Douglas [mailto:s_sourceforge@nedprod.com] >>> Sent: Dienstag, 18. März 2014 02:04 >>> To: Vicente J. Botet Escriba >>> Cc: Boris Schäling >>> Subject: Re: Welcome as organization member - HELP !!!!! >> Hi Niall, >> >>> It's possible. I am unable to accept proposals, so if you can't see >>> proposals until I accept them, that would explain it. >>> >>> I assume that only after submission closes will I be allowed to >>> accept proposals and then you can see them? Boris: you were admin >>> last year, do you know the answer? >> Vicente is a registered mentor but can't see any proposals? Unfortunately I >> don't remember anymore whether that was the case for us last year, too, >> while we were still in the student application period. Can you assign >> Vicente as a possible mentor to a proposal (the one he is interested in)? >> >> Boris >> >>> [...] > > --- > Boost C++ Libraries Google Summer of Code 2014 admin > https://svn.boost.org/trac/boost/wiki/SoC2014 > >

1 0

[GSoC 2014] Boost.odeint application
by Razvan Florea 18 Mar '14

18 Mar '14

Hello, I have added on google melange the first version of my application for Boost.odeint library. You can find it here: [1]. Please take a look on it and give me feedback in order to change the weakest parts. [1]: https://www.google-melange.com/gsoc/proposal/review/student/google/gsoc2014… Best regards, Razvan Florea

1 0

GSoC Summer Project Queries.
by Ashrith H Chinivar 18 Mar '14

18 Mar '14

Hey, This is Ashrith and I have a few queries related to the development of Boost.uBLAS library and the general expectation for the proposal with respect to the algorithmic implementation for Google SoC 2014. I hope the reply will help me narrow down on my options for this year's GSoC. The linear algebra library for Boost which expects the developer to perform Matrix multiplication and various mathematical actions w.r.t the vector spaces and matrix multiplication. Can this be done in C++ with no prior knowledge of MATLAB? I am however, very interested in contributing to the project from algorithmic point of view and including RADIX sort and various other algorithms or short-cuts that would facilitate C++. I am currently undergoing a course on Linear Algebra and Matrices. So I would e happy to consider this as the option to apply for this year's GSoC. So please reply regarding the Boost.uBLAS and other possible implementaions that is expected or which I can submit through my synopsis/proposal. I would be glad to join in on the IRC if that is possible. Thank you, Ashrith H C

1 0

GSoC 2014: boost radix sort
by Shahamat Mehdi 18 Mar '14

18 Mar '14

I have studied radix sort and it does not seem to be a very difficult task. I want to know what exactly will be required from me in this task. It will be highly appreciated if i get a link to the work that is already done so i could understand how to take it ahead. Thanks

1 0

Re: [boost] Arithmetic (opaque) type without implicit conversions
by Gonzalo BG 18 Mar '14

18 Mar '14

> What is the Arithmetic concept? Types that supports all operations provided by integral and floating point types "model" the Arithmetic "constraint" (better than the word concept in this context), as checked by std::is_arithmetic<T> [0], which is defined as std::is_integral<T> || std::is_floating_point<T>. I don't static_assert(std::is_arithmetic<T>) to allow other types that might be viable but for which there is not a specialization of std::is_integral/std::is_floating_point. [0] http://en.cppreference.com/w/cpp/types/is_arithmetic On Tue, Mar 18, 2014 at 5:38 PM, Robert Ramey <robertmacleranramey(a)gmail.com > wrote: > > > On Tuesday, March 18, 2014 8:42:02 AM UTC-7, Gonzalo BG wrote: >> >> Arithmetic<T, U = void> is a type that: >> - models the Arithmetic concept, >> > > What is the Arithmetic concept? > > >> - wraps a T that also models the arithmetic concept, and >> - can be used as a strong/opaque typedef for T. >> >> It is thus similar to Boost.Serialization's BOOST_STRONG_TYPEDEF, >> - but without any implicit conversions (only explicit ones), and >> - it supports move semantics and C++14 constexpr. >> >> The code is here: https://github.com/gnzlbg/arithmetic_type >> The Readme.md is brief, so I refer you to there for more information. >> There are also some tests, although it could use some more. >> >> The LoungeC++ and Morwen(a)codereview.stackexchange deserve credit for >> some >> improvements on the implementation. >> The remaining mistakes are only mine. >> So please, destroy my implementation if you find any mistakes. >> >> I find it useful to strongly type interfaces that involve indices, e.g. >> in >> an octree implementation where child/neighbor indices for a node with >> different types result in clear compilation errors if mixed. >> >> I hope you find it useful too. >> >> Bests, >> Gonzalo BG >> >> _______________________________________________ >> Unsubscribe & other changes: http://lists.boost.org/ >> mailman/listinfo.cgi/boost >> >

1 0

GSoC boost help
by Shahamat Mehdi 18 Mar '14

18 Mar '14

I want to work on radix sort algorithm for boost in this GSoC. Please tell me how can I contact a mentor. I have a some questions before i can submit my proposal. So, please help me.

2 1

Arithmetic (opaque) type without implicit conversions
by Gonzalo BG 18 Mar '14

18 Mar '14

Arithmetic<T, U = void> is a type that: - models the Arithmetic concept, - wraps a T that also models the arithmetic concept, and - can be used as a strong/opaque typedef for T. It is thus similar to Boost.Serialization's BOOST_STRONG_TYPEDEF, - but without any implicit conversions (only explicit ones), and - it supports move semantics and C++14 constexpr. The code is here: https://github.com/gnzlbg/arithmetic_type The Readme.md is brief, so I refer you to there for more information. There are also some tests, although it could use some more. The LoungeC++ and Morwen(a)codereview.stackexchange deserve credit for some improvements on the implementation. The remaining mistakes are only mine. So please, destroy my implementation if you find any mistakes. I find it useful to strongly type interfaces that involve indices, e.g. in an octree implementation where child/neighbor indices for a node with different types result in clear compilation errors if mixed. I hope you find it useful too. Bests, Gonzalo BG

1 0