Revert "Avoid obtaining several identical dimensions between two LSH sub-vectors by choosing orthogonal sub-vectors."

This reverts commit e85bacff7b676a7789bea6501e694cac299b21be.
This commit is contained in:
Maksim Shabunin 2014-08-08 16:59:24 +04:00
parent 032c363ea0
commit f908c2b0e3

View File

@ -346,21 +346,13 @@ inline LshTable<unsigned char>::LshTable(unsigned int feature_size, unsigned int
mask_ = std::vector<size_t>((size_t)ceil((float)(feature_size * sizeof(char)) / (float)sizeof(size_t)), 0);
// A bit brutal but fast to code
static std::vector<size_t> indices(feature_size * CHAR_BIT);
//Ensure the Nth bit will be selected only once among the different LshTables
//to avoid having two different tables with signatures sharing many dimensions/many bits
if( (indices.size() == feature_size * CHAR_BIT) || (indices.size() < key_size_) )
{
indices.resize( feature_size * CHAR_BIT );
std::vector<size_t> indices(feature_size * CHAR_BIT);
for (size_t i = 0; i < feature_size * CHAR_BIT; ++i) indices[i] = i;
std::random_shuffle(indices.begin(), indices.end());
}
// Generate a random set of order of subsignature_size_ bits
for (unsigned int i = 0; i < key_size_; ++i) {
size_t index = indices[0];
indices.erase( indices.begin() );
size_t index = indices[i];
// Set that bit in the mask
size_t divisor = CHAR_BIT * sizeof(size_t);