June 2003 - Boost-users - lists.boost.org

vector ranges
by cerenoc 10 Jun '03

10 Jun '03

Hello, I just wanted to clarify if the following behavior is correct. As far as I understood, in matrix and vector range function, the 1st argument defines the starting point and the second defines the length of the range. The following code I ran on gcc 2.96 shows that the behavior is different when starting at 0 and starting somewhere else. Is this a bug, or am I misunderstanding something? Thank you, Julian // define some vector r(3) std::cout << r << "\n\n"; boost::numeric::ublas::vector<double> rr = r(range(0, 0)); std::cout << rr << "\n\n"; rr = r(range(0, 1)); std::cout << rr << "\n\n"; rr = r(range(0, 2)); std::cout << rr << "\n\n"; rr = r(range(0, 3)); std::cout << rr << "\n\n"; //rr = r(range(1, 0)); // THIS RESULTS IN ASSERTION ERROR //std::cout << rr << "\n\n"; rr = r(range(1, 1)); std::cout << rr << "\n\n"; rr = r(range(1, 2)); std::cout << rr << "\n\n"; rr = r(range(1, 3)); std::cout << rr << "\n\n"; OUTPUT: [3](0,1,2) [0]() [1](0) [2](0,1) [3](0,1,2) [0]() [1](1) [2](1,2)

3 2

confused by matrix assignment
by cerenoc 10 Jun '03

10 Jun '03

Hi, I'm sorry I am new to this and also not up on the math notation you guys are using. I am confused by the definitions of slices and strides. I am trying to perform a reasonably simple operation of e.g. assigning a vector v(2) to a (sub)matrix m(3,3). In matlab notation, for instance: m = 0 1 2 3 4 5 6 7 8 v = 15 16 m(2, 1:2) = v Resulting in m = 0 1 2 15 16 5 6 7 8 What is the best way to do this with uBLAS? Thank you very much Julian

2 2

RE: [Boost-Users] Re: Tree Object Needed!
by Jim.Hyslop 10 Jun '03

10 Jun '03

Hossein Haeri wrote: > Does the method you mentioned work for my problem in > which there is no ordering? If yes, please explain it > in more detail (with an example). Trees are generally used with ordered sequences, aren't they? If your data are not ordered, then how do you propose to insert them into the tree? What, well, order would they appear in the tree? If you provide a little more information about the problem domain, we can help you figure out the best container to use. -- Jim

4 8

RE: [Boost-Users] Re: Tree Object Needed!
by Mark Storer 10 Jun '03

10 Jun '03

I don't mean to come off like a hard case, and I may well be wrong but... This whole line of questioning has that "homework assignment" smell to it. The initial question was _highly_ school-related: > I need to have some template doing the jobs of an > ordinary BST (Binary Search Tree). Nothing that fancy, > but something at the level we all have been educated > in CS1/FCS1, or similar courses. Does boosts > repository contain such an object? The utter lack of background detail leads me to believe that this is just a toy problem... a homework assignment. And then there's the yahoo email address... most proffesional programmers have @mybusiness addresses, though that doesn't mean everyone here is a proffesional. Oh, and the final bit of oddness: > Why don't I write it myself? Simply, because I want to > optimise the time! How is waiting a week or so for some mailing list to hand you an answer "optimizing the time"? This has just been bothering me, I felt the need to speak out. If I'm wrong, then I offer my heartfelt appoligies. This may be what Mr. Hyslop was trying to dig at, I'm just not a patient or subtle person. And I don't mind appologizing for my mistakes (lots of practice, as you might imaging with this sort of behavior). --Mark Storer Software Engineer Cardiff Software #include <disclaimer> typdef std::disclaimer<Cardiff> Discard;

2 1

Threads Library
by Brian Plummer 09 Jun '03

09 Jun '03

Atomic operations (atomic_add, atomic_subtract etc) are not handled in the Boost Threads library. My guess is that this is because this cannot be accomplished, based on what I know about this topic (which isn't too much), in a portable way. Is this the reason? If not, having these operations in the Threads library would be a big benefit. Thanks, Brian

2 1

counted_base vs. sp_counted_base - boost 1.30.0 functionality deprecated
by Anders Hybertz 09 Jun '03

09 Jun '03

Hi Up to and including version 1.29.0 of the boost library I had an object which inherited from boost::counted_base, and could actually within that object control when it should be deleted. The object spawned a thread which was listening on a queue, when there was pushed an shutdown message on the queue, I could safely delete the object. In boost 1.30.0 the counted_base changed name and namespace to boost::detail::sp_counted_base, but the way it works also changed - so now my object's destructor is called whenever it's going out of scope - which I don't want. Any ideas - or should I provide a small sandbox example Thanks in advance Anders Hybertz

3 4

[crc] Simple example please
by Alex Henderson 09 Jun '03

09 Jun '03

Hi, I'd like to calculate a value for a cyclic redundancy check (crc) for a number of files. This isn't going to be transmitted anywhere, I simply want to compare the values to see whether the files are exactly the same. I know I can simply check byte for byte, but I've got a few thousand files and it's easier to calculate the crc for each and see whether any are the same first. I've looked at the crc library, but can't get my head round the example. It seems to test the library rather than show how its used. I'm after something like std::vector<char> MyStream1; // the whole file (about 500kb) boost::SomeCRCTypeThatIsOptimisedForSafetyAndLargeishBlocks<???> x; long = x.checkme(MyStream1); Could someone please point me to some code where a simple example is detailed? Thanks, Alex

1 0

confused by ublas submatrix assignment
by cerenoc 08 Jun '03

08 Jun '03

Hi, I am new to this and am confused by the (sub)matrix operations and also not up on the math nomenclature this group is using such as slices and strides. I am trying a reasonably simple operation of assigning a vector v(2) to a submatrix (3,3). For example, in matlab notation: m = 0 1 2 3 4 5 6 7 8 v = 15 16 m(2, 1:2) = v resulting in: m = 0 1 2 15 16 5 6 7 8 What is the best way to do this with ublas? I suppose I need to convert the vector into another type first, but which one - a slice, a range? I tried both with no success. Thanks a lot, Julian

2 1

uBLA performance
by Matt Hurd 07 Jun '03

07 Jun '03

Hi, Ran off a simple test of the uBLA performance for sqare matrix multiplication. Looks quite good. Put in Intels Math Kernel Library for good measure using uBLA as the data holder but with a call to dgemm. Do you think these are the kind of numbers I should be expecting? Numbers below. Added a chart to the files section http://groups.yahoo.com/group/Boost-Users/files/ called uBLA Performance.pdf Used head of CVS from today. Regards, Matthew Hurd ___________________________________________ Figures in the raw: Matrix Size intel7.1 vc7.0 mingw intel_mkl6.0_vc7.0 ___________________________________________ 1 2 2 2 1 2 13 12 13 11 3 32 30 34 31 4 51 46 59 70 5 67 59 85 109 7 70 78 92 180 9 85 94 113 262 12 99 115 139 304 16 110 131 159 362 20 118 127 178 355 26 125 135 196 334 33 132 148 209 390 42 134 158 217 389 53 130 163 228 451 67 133 171 233 459 84 138 177 241 467 105 133 176 239 473 132 133 178 241 468 166 106 174 233 448 208 90 148 196 445 261 40 63 63 427 327 24 43 45 434 409 24 44 45 421 512 11 12 12 404 640 12 12 12 459 800 36 38 38 455 1000 36 37 38 448 ___________________________________________

1 0

Newbie query: no "deamon" thread?
by dick.bridges＠tais.com 07 Jun '03

07 Jun '03

I need to spawn worker threads that simply perform a task and exit. All the examples I can find seem to use join() to cleanup the thread resources from another thread. Isn't there some way to declare worker threads so that they "do the right thing" as they exit the thread function? TIA

3 4