January 2009 - Boost-users - lists.preview.boost.org

Help with a generic algorithm using boost::MultiArray/Multiple nested range traversal
by Jesse Perla 12 Jan '09

12 Jan '09

Steven: >int index_of(const multi_array<int, 3>& m, int* iter, int dimension) { > int offset = iter - m.origin(); > return(offset / m.strides()[dimension] % m.shape()[dimension] + >m.index_bases()[dimension]); } Thanks for the code suggestion. Does this code work for N dimensions? it appears to. I would prefer to keep a tight count of the indices in the iterator itself to ensure that we don't need to do this expensive calculation each time, but this looks like a pretty good starting point and would work similarly for my sequential ublas::matrix iteration as well. -Jesse

1 0

Help with a generic algorithm using boost::MultiArray/Multiple nested range traversal
by Jesse Perla 12 Jan '09

12 Jan '09

Hi Rob, Thanks for the response. I also think this is a pretty important use case for this great library. Let me justify why I think an Iterator based approach is the best: * For maximum efficiency (I will be using this in numerical simulations), we will want have an incrementing pointer within the data structure pointing to the actual object, and not just the iterator. * Also, if we have a single forward iterator, then we should be able to reuse a bunch of standard algorithms in the cases where we do not need the actual index values or the order of execution doesn't matter. * In the iterator algorithm, I would love to check the ordering of the data from the template parameters. If it is fortran ordering then we would do the loops in the opposite order for example, and this should be based on querying the data structure in the iterator code. * As for your range based use case, I don't know enough about how iterators and ranges interact to see how that would work. * As to implementation of this, thanks for the pointer to Fusion. This indeed seems to be the best approach since the loops can be generated at compile time, and compilers could then do some very fancy optimization on them. Sadly, I am not a template metaprogrammer, or a very good C++ programmer in general, and am terrified to learn it and lose sight of my non-CS research. Now for the semantics of this, it seems to me that the way the iterator concepts are designed, the distance() function is intended to pass back the related index information. Now my question is whether it is OK for the semantics of distance to pass back a boost::array<NUMDIMS> instead? Seems reasonable to me that it would not be based on passing back a scalar value in this case because that is arbitrarily dependent on the column vs. row ordering. And we can easily generate a boost::array from the distance function by looking at the distance between the .begin() for all of our current positions in the sub-iterators? If the iterator framework was there and the distance implemented as vectors, I can imagine the code looking like: int n1 = 3; int n2 = 4; int n3 = 5; boost::multi_array<double, 3> A(boost::extents[n1][n2][n3]); //NOTE SURE OF THE RIGHT SEMANTICS FOR GETTING AN ITERATOR OUTSIDE OF ONE GENERATED BY THE CONTAINER. multi_array_iterator<double, 3> begin = mult_sequential_iterator_begin(A); multi_array_iterator<double, 3> end = mult_sequential_iterator_end(A); for(mult_sequential_iterator<double, 3> iter = begin; iter != end; ++iter) *iter = f(distance(begin, iter); //POSSIBLE IMPLEMENTATION OF DISTANCE template<int NUMDIMS> boost::array<NUMDIMS> distance(const mult_sequential_iterator& iter1, const mult_sequential_iterator& iter1) const { static_assert(iter1::dimensionality == iter2::dimensionality) boost::array<NUMDIMS> dist; foreach dimension //PSEUDOCODE, FROM FUSION???? array<dimension> = distance(iter1.get_sub_iter<dimension>(), iter2.get_position<dimension>()); //here get_sub_iter returns the sub-iterator stored for that current dimension in the iterator. } template<int NUMDIMS> //HERE THE FUNCTION WOULD BE BASED ON A BOOST::ARRAY. double f(const boost::array<NUMDIMS>& indices) { ...} //AND LETS SAY WE WANTED TO NEGATE THE ENTIRE DATA STRUCTURE, NOW CAN USE STD: int n1 = 3; int n2 = 4; int n3 = 5; boost::multi_array<double, 3> A(boost::extents[n1][n2][n3]); boost::multi_array<double, 3> B(boost::extents[n1][n2][n3]); std::transform(mult_sequential_iterator_begin(A), mult_sequential_iterator_end(A), mult_sequential_iterator_begin(B), negate<double>()); //AND IF WE ADDED THIS ITERATOR TO THE CODE FOR THE MULTDIM: A.begin_sequential() could return the sequential iterator...

2 1

what are the different problems that can occur in network connections and send/recive, what are the related exceptions & errors, and what are the solutions (boost.asio)?
by Saeed Aliakbarian 12 Jan '09

12 Jan '09

While working with boost.asio networking features, in many cases problems may occur due to network connection erros, programmin errors, etc. In each case a different event occurs in the program. Most of the times, and error code is returned or an exception is thrown. I could not find a list of this error causes and their corresponding error codes or exceptions. The only way I could mange only a part of this problem was trough testing, creating a specific error manually, catching the exception or the error code, and the try to manage it in the correct way I suggested myself!!!! If there is not such a page containing my required information, do you agree there is a need for one? Can anyone help me find a better way to do what I am trying to do? Regards Saeed Aliakbarian

1 0

Boost Units standardization?
by Mark Volkmann 12 Jan '09

12 Jan '09

What is the status on getting the Units library to be part of the C++ standard? Is this being pursued? --- Mark Volkmann

2 1

[date_time]utc_to_local crashes
by Jan Deinhard 12 Jan '09

12 Jan '09

Hi, I get a crash calling utc_to_local with dates past 2038-Jan-19 03:14:07. I'm compiling with Visual C++ 2005. using namespace boost; posix_time::ptime pt(gregorian::date(2038, 1, 19), posix_time::hours(3) + posix_time::minutes(14) + posix_time::seconds(8)); date_time::c_local_adjustor<posix_time::ptime>::utc_to_local(pt); Jan

1 0

[Ranges] make_view: when will it be part of a release?
by Christoph Duelli 12 Jan '09

12 Jan '09

I recently had the need to produce from a vector<uint> a string representation (like for IPs) like "23.1.4.742.234.0". I thought that if only those were uints were strings I could boost::join them together. After a bit of searching the internet I found: make_view... very helpful, as indeed operating on iterators and containers only is a bit restrictive. So, first, I'd like to express my thanks for providing such helpful libraries (I love it how Boost.Bind and ASIO fit together, e.g.!). Great job everyone! Is there a better way to get to the string representation hinted to above? What I did is: #include <boost/view.hpp> // not yet part of Boost 1.37 // Helper to convert an uint to a std::string struct Stringer { std::string operator() (unsigned u) const { return boost::lexical_cast<std::string>(u); } }; boost::function<std::string(unsigned)> stringer = Stringer(); // way better than writing the loop by hand! std::string s = boost::join(boost::make_view(data_, stringer), "."); // Does not compile with gcc 4.2.1 // std::string s = boost::join(boost::make_view(data_, Stringer()), "."); Is there some existing class that makes my helper class Stringer unnecessary? (Can I use lexical_cast here w/o a helper class?) Is it possible to avoid putting the functor into a local Boost.Function variable? When will range_ex and make_view() be part of a Boost release? 1.38? (Please?) Best regards, and thank you -- Christoph Duelli

5 5

binary archive with text file
by gtsml owevwr 12 Jan '09

12 Jan '09

Hi, I just spend a entire day to realize this code was wrong: std::ofstream ofs(filename.c_str()); boost::archive::binary_oarchive oa(ofs); oa << (*this); And this one was right: std::ofstream ofs(filename.c_str(), std::ios::binary); boost::archive::binary_oarchive oa(ofs); oa << (*this); The consequence of the first code is a "stream_error" throw which doesn't say much about the problem. Instead, an exception should be thrown earlier on if the file is not in binary mode. thanks alot to all the boost contributors, Frank

2 2

Re: [Boost-users] Serialization of forward declared class
by niranjan bangera 12 Jan '09

12 Jan '09

hi all, > i have done in the following way.. even am getting some errors, did i miss > any of the headers, or added extra(added extra for sure). > > Any suggestions will be accepted > > #include <boost/serialization/utility.hpp> > #include <boost/serialization/collection_traits.hpp> > #include <boost/serialization/serialization.hpp> > #include <boost/serialization/base_object.hpp> > #include <fstream> > #include <boost/archive/text_iarchive.hpp> > #include <boost/archive/text_oarchive.hpp> > #include <boost/serialization/assume_abstract.hpp> > > class fd; > > class base: class derived > { > ..... > fd *ptr; > ..... > > friend class boost::serialization::access; > template<class Archive> > void serialize(Archive & ar, const unsigned int file_version) > { > (void) file_version; > ar & ptr; > ar & boost::serialization::base_object<derived>(*this); > > } > > }; //end of class base > > Version::1.36.0 > compliler: g++ 4.1.1.... > > i am getting a error : > error: forward declaration of 'struct fd' > /boost_1_36_0/include/boost-1_36/boost/serialization/access.hpp:109: error: > inva > lid use of undefined type 'struct > > /project/wipro_virtutech_nobackup/users/niranjan/project_new/boost_1_36_0/include/boost-1_36/boost/serialization/extended_type_info_typeid.hpp:117: > note: neither the destructor nor the class-specific operator delete will be > called, even if they are declared when the class is defined. > > Thanks in advance..... >

2 1

Help with a generic algorithm using boost::MultiArray
by Jesse Perla 11 Jan '09

11 Jan '09

Hi:I would like to write an algorithm that operates on a multi-array and am having trouble figuring out how to work with arbitrary numbers of dimensions. A simplified version of my problem that captures all of the types of access I need is: 1) You have a multi-array of (statically defined at compilation) dimensionality N. 2) This data is filled in, but the number of elements in each dimension will frequently be different. 3) Write a function that takes a reference to this multiarray and assigns it the value which is the sum of all of the actual indices(this is contrived, but if I can solve this, I can solve my problem... The key is I need an iterator moving through all dimensions, and I need the index numbers for my calculations). Here is some pseudocode that gives the kind of iterator I would love to have: template<int N> void test_setting(multi_array<double, N>& arr) { //HERE IS PSEUDOCODE: LETS SAY THERE WAS A WAY TO GET AN INDEX THAT YIELDED THE ACTUAL INDEX NUMBERS BUT STILL ACTED AS A POINTER. for(FANCYINDEX = arr.begin(); FANCYINDEX != arr.end(); ++FANCYINDEX) { boost::array<int, N> indicies = FANCYINDEX.get_indices(); //SAY IT RETURNED THE ACTUAL NUMBERS *FANCYINDEX = sum(indicies); //Alternatively, could we go: arr(indicies) = sum(indices); if there is no way for it to be a proper pointer. } } //TO CALL: const int NUMDIMS = 3; //Static at compile time. int n1 = 3; //These might be dynamic. int n2 = 4; int n3 = 5; boost::multi_array<double, NUMDIMS> A(boost::extents[n1][n2][n3]); test_setting(A); Any ideas on how this could be implemented without too much code? Thanks, Jesse

2 1

[proto] iterating array expressions
by Daniel Oberhoff 11 Jan '09

11 Jan '09

Hi, First off: much thanks for previous replies to related problems. One comment: I got told to properly use proto on my array class I needed to separate data from user interface. I found now that it is possible to define expressions non-intrusively without having to change my array structure, but still I understand the motivation, and possibly this way is better anyway. Anhow, now to my new problems. I have now set up proto to be able to evaluate expressions such as: A(_i,_j) + B(_j,_i*_j-2) at a given index (i.e. substituting given numbers for index placeholders _i and _j). Here A and B are of my own n-dimensional array class. Now I would like to iterate such expressions efficiently to evaluate assignment expressions like: C = A(_i,_j) + B(_j,_i*_j-2); efficiently. What I want to do is base it around basic "iterators" for each array indexing expression ( i.e. B(_j,_i*_j-2) and A(_i,_j) in the example) that know how to move to a certain setting for the index placeholders and how to advance along the index placeholders (e.g. advance the data pointer for _i -> _i+1, _j -> _j+1). I think I could implement such iterators (not in the STL sense, since I dont use begin/end, but enough for what I need) for the array indexing expressions, but I dont know how to generate the iterator for the complete expression. I would like such an iterator be generated automatically given the individual iterators for the array indexing expressions and be able to call move, advance, and evaluate on such an iterator. Any hints on where to start before I lose myself deeply in the documents and examples? Any help much appreciated. Best Daniel

2 1