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

NDEBUG, affects floating point math?
by Stephen Crowley 28 Feb '03

28 Feb '03

I'm getting some seriously strange problems with regard to numerical accuracy when computing matrix products that depends on whether or not NDEBUG is defined when I compile. I wrote a small piece of C code to calculate a symmetric-matrix by row-matrix product by directly accessing the memory (which btw is about 1000% faster than ublas's generated code with gcc 3.2.3 and -O3, why is ublas so slow?) And when I compile with -DNDEBUG the results differ by around +/- 1e-16, but when its not defined the results match almost perfectly. Is there something strange going on with regard to FPU control words or types or anything? I'm using type double btw. If more info is needed I can post a test case, but I thought it'd make a quick inquery to see if there is a fast answer. -- Stephen -------- This message was scanned for viruses by Digital Passage.

1 0

Re: [Boost-Users] Problem with time_period
by Jeff Garland 28 Feb '03

28 Feb '03

Alex wrote: >... start omitted ... >time_period p(ptime(date(2003, 03, 25)), ptime(date(2003, 04, >25))); >date_duration duration(days(1)); >p.shift(duration); >---------------------------------------------------- I'm not sure where 'days' is coming from -- its not a class in the library. Based on the compiler error it looks like the issue is using date_duration intead of time duration with the time_period. So instead you could do something like: hours duration(24); Or if you really want a class called days you can do something like: class days : public boost::posix_time::time_duration { public: days(int day_count) : boost::posix_time::time_duration(day_count*24) {} }; HTH. And, BTW in the upcoming 1.30 release you will be able to get rid of all the DATE_TIME_INLINE stuff... Jeff

2 2

mpl::if_ or mpl::apply_if?
by Markus Werle 27 Feb '03

27 Feb '03

Hi! I am trying to migrate some parts of my code from loki to boost::mpl. Q1: which is the best way to perform nested if-then-else? Assume given "values": bool B1, bool B2; class ff, class tf, class ft, class tt; I replaced ----------------------------------------------------------- typedef typename Loki::Select<B1, typename Loki::Select<B2, tt, tf>::Result, typename Loki::Select<B2, ft, ff>::Result >::Result type; ----------------------------------------------------------- with typedef typename apply_if < bool_c<B1>, apply_if < bool_c<B2>, identity<tt>, identity<tf> >, apply_if < bool_c<B2>, identity<ft>, identity<ff> > >::type type; Q2: Is apply_if always to be preferred over if_? Q3: (OT) Is there a mpl-equivalent to run-time case? best regards, Markus

4 10

Re: [Boost-Users] Re: are namespaces of boost libs required to be lowercase?
by Beman Dawes 27 Feb '03

27 Feb '03

At 11:37 AM 2/27/2003, William E. Kempf wrote: > >Markus Werle said: >> A hard burden You lay on developers. > >It's not a burden. > >> So maybe the code has to be piped through a >> code de-beautifier when it comes to a boostification. > >That's a bit harsh. If *you* don't care for the style, you're welcome to >that opinion, but voicing it in this manner isn't going to help anyone to >agree with you. > >> IMHO (UsingThisConventionHere == CommonPracticeElsewhere) > >It's not common practice everywhere (I know of as many places that follow >the C++ standard naming conventions as those that follow what I'll call >"Java" naming conventions). But the important thing is that the whole >point of Boost is to be a test bed and development area for libraries that >*might* be considered for inclusion in the standard. As such, we have to >follow the standard naming conventions. Exactly. And it isn't just a theoretical "*might* be considered for inclusion in the standard". A bunch of the Boost libraries *are* under consideration by the Standards Committee right now, and more will be in the future. For a library to be considered seriously for standardization, it helps a great deal if every aspect, including names, has been subject to prior use. That flushes out issues that otherwise would be hidden. So there is little chance Boost will change naming conventions. I suppose we might accept a library which used a different set of conventions if there was some special rationale. But I can't imagine what that rationale would be. It certainly would not be just that a developer prefers another convention. --Beman

1 0

boost 1.29.0: shared_from_this with const argument
by Michael Brinkmann 27 Feb '03

27 Feb '03

Hi, In my opinion, the call "shared_from_this(T* p)" in the file "boost/shared_ptr.hpp" is meant to be "shared_from_this(T const* p)" Am I right? kind regards Michael

2 1

UBLAS: advice on argument types please
by speedsnaii <speedsnaii＠yahoo.de> 27 Feb '03

27 Feb '03

I am trying to find a good choice for argument types to a function. (Please be patient, I am a poor user of C++ who has not yet understood all this expression template magic.) I'd like to have a function: void foo( whattype arg) { } matrix<double> m(3,3); vector<double> v(6); and call it like: foo(m.column(0)); foo(m.row(0)); foo(v); Is this feasible? Should I use iterators for whattype? (how do I access size information then?) Am I trying to misuse the ublas when asking such a question? (Which container should I use in this case?) In the hope anyone can shade some light on it. Thank you, Roland

2 11

Building with Sun Workshop 6
by markrday <markrday＠yahoo.com> 27 Feb '03

27 Feb '03

I've searched the archives, and haven't found a report of successfully building Boost using the Sun Workshop compiler, though some messages imply that people are using at least some of the libraries with Workshop. Although I realize we could build the libraries with gcc, we have quite a few projects set up in the workshop environment, and I'd like to have the option to use Boost libraries with them. Obviously nobody has built an appropriate toolset. In my naive attempts to create a toolset definiton, and in trying to create a static library from components like Date-Time library, I get error messages like: add_const.hpp, line 34: Warning: Too few arguments in macro BOOST_PP_CHECK_2. There are also errors involving BOOST_PP_LIST_FOR_EACH_I_M (among others). Can anyone offer me hints on how to get started in resolving these? Ultimately I'd like to have the whole suite compiled, but for the time being I'd be happy just using Date-Time. I'm willing to build a toolset file for this compiler, but I don't seem to have the expertise. If there's anyone that is able to provide guidance, I can put some time into this, and submit a toolset definition file so others can use Boost with this compiler.

3 3

Problem with time_period
by Alejandro Escalante Medina <alejandro_escalante_medina＠yahoo.com.mx> 27 Feb '03

27 Feb '03

The following simple code doesn't compile: ---------------------------------------------- #define DATE_TIME_INLINE #include <boost/date_time/gregorian/gregorian.hpp> #include <boost/date_time/posix_time/posix_time.hpp> using namespace boost; using namespace boost::posix_time; using namespace boost::gregorian; ... time_period p(ptime(date(2003, 03, 25)), ptime(date(2003, 04, 25))); date_duration duration(days(1)); p.shift(duration); ---------------------------------------------------- the compiler (VC++ 7.0) gives the following: error C2664: 'boost::date_time::period<class boost::posix_time::ptime,class boost::posix_time::time_duration>::shift' : cannot convert parameter 1 from 'boost::gregorian::date_duration' to 'const boost::posix_time::time_duration &' Reason: cannot convert from 'boost::gregorian::date_duration' to 'const boost::posix_time::time_duration' No constructor could take the source type, or constructor overload resolution was ambiguous I think this should work out of the box, since I've been doing lots of other things with date_time and they work fine. Any ideas? Thanks in advance, Alex

1 0

uBLAS - Matrix Inversion
by Thomas Reisinger 25 Feb '03

25 Feb '03

Dear Newsgroup, I'm new in the field of numeric computing and I would like to learn more about it. To solve my problem I need to invert a matix and use the intermediate result as the input for following calculations. I browsed the uBLAS library and played a little bit with it but I didn't find a routine for matrix inversion. Is there any possibility to preform it in a single step? Is this package the wrong tool for doing this and I should use LAPACK++ or similar stuff? Thanks in advance, Thomas -- Thomas Reisinger Braunschweig Germany mailto:t.reisinger@tu-bs.de

4 3

date_time problems
by Eric Crampton 25 Feb '03

25 Feb '03

I'm trying to test the date_time functionality of boost and am having some difficulties. I would like to simply get the current local time into a ptime and write it out as an ISO-8601 string. Using Boost 1.29 and g++ 3.2, when I compile the following code I get a number of *linker* errors. Code: #include <boost/date_time/posix_time/posix_time_types.hpp> #include <boost/date_time/posix_time/time_formatters.hpp> int main() { boost::posix_time::ptime pt = boost::posix_time::second_clock::local_time(); std::cout << boost::posix_time::to_iso_extended_string(pt) << std::endl; } That code seems pretty straightforward to me. The linker errors look like the following: test.cpp:7: undefined reference to `boost::date_time::gregorian_calendar_base<boost::date_time::year_month_day_base<boost::gregorian::greg_year, boost::gregorian::greg_month, boost::gregorian::greg_day>, unsigned long>::day_number(boost::date_time::year_month_day_base<boost::gregorian::greg_year, boost::gregorian::greg_month, boost::gregorian::greg_day> const&)' test.cpp:6: undefined reference to `boost::date_time::gregorian_calendar_base<boost::date_time::year_month_day_base<boost::gregorian::greg_year, boost::gregorian::greg_month, boost::gregorian::greg_day>, unsigned long>::from_day_number(unsigned long)' (snipped...) Best as I can tell (cd $BOOST_ROOT/boost/date_time && find . -name '*.cpp' -print), there are no .cpp files to be compiled... Anyone know what's up? --Eric _________________________________________________________________ Protect your PC - get McAfee.com VirusScan Online http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

4 3