Snippets for EMA-course in Yandex Data Analysis School

main.cpp

#include <stxxl/vector>
#include <stxxl/random_shuffle>

#include <boost/random/uniform_int.hpp>
#include <boost/random/uniform_real.hpp>
#include <boost/random/variate_generator.hpp>
#include <boost/generator_iterator.hpp>

class Profiler
{
private:
    stxxl::stats_data StatsBefore;
    stxxl::stats_data StatsAfter;

public:    
    Profiler()
    {
        StatsBefore = *stxxl::stats::get_instance();
    }

    ~Profiler()
    {
        StatsAfter = *stxxl::stats::get_instance();

        Dump();
    }

private:
    void Dump()
    {
        std::cerr << (StatsAfter - StatsBefore) << std::endl;
    }
};

int main(int argc, const char** argv)
{
    try {
        (void)stxxl::block_manager::get_instance();
    } catch (const std::exception& ex) {
        std::cerr << "Error while initializing STXXL: " << ex.what();
    }

    typedef stxxl::VECTOR_GENERATOR<
        float,
        1, // Number of blocks in a page
        1, // Number of pages kept in memory
        2 * 1024 * 1024, // Block size
        stxxl::RC, // Allocation strategy
        stxxl::lru>::result FloatVector;

    unsigned int M = 32 * 1024 * 1024;
    unsigned int Nb= 64 * 1024 * 1024;
    unsigned int N = Nb / sizeof(float);

    FloatVector v(N);

    std::cerr << "M(b) = " << M << " ; N(b) = " << Nb << " ; N = " << N << std::endl;

    boost::uniform_real<> uniformDistribution(0,1);
    boost::mt19937 generator(42);
    boost::variate_generator< boost::mt19937&, boost::uniform_real<> > uniform(generator, uniformDistribution);

    {
        std::cerr << " * Generating data..." << std::endl;;
        for (size_t i = 0; i < v.size(); ++i) {
            v[i] = uniform();
        }
    }

    {
        Profiler profiler;
        std::cerr << " * Shuffling data..." << std::endl;;
        stxxl::random_shuffle(v.begin(), v.end(), M);
    }

    return 0;
}

Makefile

STXXL_ROOT      ?= ../stxxl
STXXL_CONFIG    ?= stxxl_boost.mk

include $(STXXL_ROOT)/$(STXXL_CONFIG)

CXX              = $(STXXL_CXX)
CPPFLAGS        += $(STXXL_CPPFLAGS)
LDLIBS          += $(STXXL_LDLIBS)

CPPFLAGS        += -O3 -Wall -g 

my.bin: main.o
	$(CXX) $(CXXFLAGS) $(CPPFLAGS) $(LDFLAGS) main.o -o $@ $(LDLIBS)