void CompactTriples::load(ModifiableTriples &triples, ProgressListener *listener) {
triples.sort(order);
IteratorTripleID *it = triples.searchAll();
vector<unsigned int> vectorY, vectorZ;
unsigned int lastX, lastY, lastZ;
unsigned int x, y, z;
// First triple
if(it->hasNext()) {
TripleID *triple = it->next();
swapComponentOrder(triple, SPO, order);
lastX = x = triple->getSubject();
lastY = y = triple->getPredicate();
lastZ = z = triple->getObject();
vectorY.push_back(y);
vectorZ.push_back(z);
numTriples++;
}
// Rest of the triples
while(it->hasNext()) {
TripleID *triple = it->next();
//cout << "111> " << triple << endl;
swapComponentOrder(triple, SPO, order);
//cout << "222> " << triple << endl;
x = triple->getSubject();
y = triple->getPredicate();
z = triple->getObject();
if(x!=lastX) {
vectorY.push_back(0);
vectorY.push_back(y);
vectorZ.push_back(0);
vectorZ.push_back(z);
} else if(y!=lastY) {
vectorY.push_back(y);
vectorZ.push_back(0);
vectorZ.push_back(z);
} else {
vectorZ.push_back(z);
}
lastX = x;
lastY = y;
lastZ = z;
NOTIFYCOND(listener, "Converting to CompactTriples.", numTriples, triples.getNumberOfElements());
numTriples++;
}
delete it;
VectorUIntIterator itY(vectorY);
VectorUIntIterator itZ(vectorZ);
streamY->add(itY);
streamZ->add(itZ);
#if 0
// Debug Adjacency Lists
cout << "Y" << vectorY.size() << "): ";
for(unsigned int i=0;i<arrayY->getNumberOfElements();i++){
cout << arrayY->get(i) << " ";
}
cout << endl;
cout << "Z" << vectorZ.size() << "): ";
for(unsigned int i=0;i<arrayZ->getNumberOfElements();i++){
cout << arrayZ->get(i) << " ";
}
cout << endl;
#endif
}
ERMsg CKNearestNeighbor::LoadFromString(const string& dataX, const string& dataY)
{
ERMsg msg;
vector<string> species = GetSpeciesList(dataX);
if (species.empty())
{
msg.ajoute("Invalid input file. Need to have one column named SpeciesID");
return msg;
}
std::stringstream streamX(dataX);
std::stringstream streamY(dataY);
CSVIterator itX(streamX);
CSVIterator itY(streamY);
CSVRow const& headerX = itX.Header();
CSVRow const& headerY = itY.Header();
set<size_t> Xset = headerX.FindAll("X_");
set<size_t> Yset = headerY.FindAll("Y_");
vector<size_t> Xpos(Xset.begin(), Xset.end());
vector<size_t> Ypos(Yset.begin(), Yset.end());
vector<size_t> coordinatePos;
if (Xpos.empty())
{
msg.ajoute("Invalid input file. Doesn't contain X value with prefix: X_");
return msg;
}
m_X.SetNbDimension(Xpos.size()*species.size());
for (size_t s = 0; s < species.size(); s++)
for (size_t i = 0; i < Xpos.size(); i++)
m_X.SetDimensionName(s*Xpos.size() + i, headerX[Xpos[i]] + "_" + species[s]);
m_Y.resize(Ypos.size());
for (size_t i = 0; i < Ypos.size(); i++)
m_Y[i].SetName(headerY[Ypos[i]]);
for (int x = 0; itX != CSVIterator(); x++)
{
m_X.resize(m_X.size() + 1);
m_X[x].SetNbDimension(Xpos.size()*species.size());
for (size_t s = 0; s < species.size(); s++, ++itX)
{
ASSERT(itX != CSVIterator());
CSVRow const& line = *itX;
for (size_t i = 0; i < Xpos.size(); i++)
{
if (Xpos[i] < (int)line.size())
{
if (line[Xpos[i]] != "NA")
{
m_X[x][s*Xpos.size() + i] = ToFloat(line[Xpos[i]]);
}
else
{
msg.ajoute("Reading error at line : " + ToString(i + 1));
msg.ajoute("Row with missing values (\"NA\") not allowed.");
return msg;
}
}
else
{
msg.ajoute("Reading error at line : " + ToString(i + 1));
msg.ajoute("Expected number of columns : " + ToString(Ypos.size()));
msg.ajoute("Number of columns find : " + ToString(line.size()));
return msg;
}
}
}
}
//for(size_t i=0; i<Ypos.size(); i++)
for (; itY != CSVIterator(); ++itY)
{
CSVRow const& line = *itY;
for (size_t i = 0; i < Ypos.size(); i++)
{
if (Ypos[i] < (int)line.size())
{
if (line[Ypos[i]] != "NA")
{
m_Y[i].push_back(ToFloat(line[Ypos[i]]));
}
else
{
msg.ajoute("Reading error at line : " + ToString(i + 1));
msg.ajoute("Row with missing values (\"NA\") not allowed.");
return msg;
}
}
else
{
msg.ajoute("Reading error at line : " + ToString(i + 1));
msg.ajoute("Expected number of columns : " + ToString(Ypos.size()));
msg.ajoute("Number of columns find : " + ToString(line.size()));
return msg;
}
}
}
return msg;
}
