static void
loadmult_datafile_sub(istream &f, bool binary, const char *fname,
xvec_t &xp, yvec_t &yp, int &maxdim, int maxrows)
{
cout << "# Reading file " << fname << endl;
if (! f.good())
assertfail("Cannot open " << fname);
int pcount = 0;
while (f.good() && maxrows--)
{
double y;
SVector x;
y = (f.get());
x.load(f);
if (f.good())
{
xp.push_back(x);
yp.push_back(y);
pcount += 1;
if (x.size() > maxdim)
maxdim = x.size();
}
}
cout << "# Read " << pcount << " examples." << endl;
}
void
load(const char *fname, xvec_t &xp, yvec_t &yp)
{
cout << "Loading " << fname << "." << endl;
igzstream f;
f.open(fname);
if (! f.good())
{
cerr << "ERROR: cannot open " << fname << "." << endl;
exit(10);
}
int pcount = 0;
int ncount = 0;
bool binary;
string suffix = fname;
if (suffix.size() >= 7)
suffix = suffix.substr(suffix.size() - 7);
if (suffix == ".dat.gz")
binary = false;
else if (suffix == ".bin.gz")
binary = true;
else
{
cerr << "ERROR: filename should end with .bin.gz or .dat.gz" << endl;
exit(10);
}
while (f.good())
{
SVector x;
double y;
if (binary)
{
y = (f.get()) ? +1 : -1;
x.load(f);
}
else
{
f >> y >> x;
}
if (f.good())
{
assert(y == +1 || y == -1);
xp.push_back(x);
yp.push_back(y);
if (y > 0)
pcount += 1;
else
ncount += 1;
if (x.size() > dim)
dim = x.size();
}
if (trainsize > 0 && xp.size() > (unsigned int)trainsize)
break;
}
cout << "Read " << pcount << "+" << ncount
<< "=" << pcount + ncount << " examples." << endl;
}
void SVSet::deleteVec(DLPSV* ps)
{
/* delete last entries, in an SVECTOR and also in an DLPSV there is always the position -1 used for memorizing
* the size; this is the reason why we need to delete max()+1 entries
*/
if (list.last() == ps)
removeLast(ps->max() + 1);
/* merge space of predecessor with position which will be deleted, therefore we do not need to delete any
* memory or do an expensive memmove
*
* @note an SVECTOR and also in an DLPSV memorize the size always on position -1; this is the reason why we
* need to set the new mem size to the old combined size + 2
*/
else if (list.first() != ps)
{
SVector* prev = ps->prev();
int sz = prev->size();
prev->setMem(prev->max() + ps->max() + 2, prev->mem());
prev->set_size(sz);
}
/* delete the front entries of the first list entry and correct the memory pointers in the vectors */
/* @note we do this by merging the first both vectors, move the entries from the second vector up front, and
* correct the size
*/
else
{
SVector* next = ps->next();
int sz = next->size();
int bothmax = next->max() + ps->max();
int offset = 0;
/* the first element does not need to start at the beginning of the data array; why ??? */
while( &(this->SVSetBase::operator[](offset)) != ps->mem() )
{
++offset;
assert(offset < size());
}
/* move all entries of the second vector to the front */
for(int j = 0; j <= sz; ++j)
{
this->SVSetBase::operator[](offset + j) = next->mem()[j];
}
/* correct the data memmory pointer and the maximal space */
next->setMem(bothmax + 2, ps->mem());
/* correct size */
next->set_size(sz);
}
list.remove(ps);
}
DSVector::DSVector(const SVector& old)
: theelem( 0 )
{
allocMem(old.size() + 1);
SVector::operator= ( old );
assert(DSVector::isConsistent());
}
Real Vector::operator*(const SVector& v) const
{
assert(dim() >= v.dim());
int i;
Real x = 0;
for (i = v.size(); i-- > 0;)
x += val[v.index(i)] * v.value(i);
return x;
}
void
load_classes(const char *fname, int &nclass)
{
string filename = fname;
ifstream f;
SVector x;
f.open(fname);
x.load(f);
nclass = x.size();
}
void SVSet::xtend(SVector& svec, int newmax)
{
if (svec.max() < newmax)
{
if (possiblyUnusedMem * memFactor > memSize())
memPack();
assert(has(&svec));
DLPSV* ps = static_cast<DLPSV*>( & svec );
if (ps == list.last())
{
int sz = ps->size();
ensureMem (newmax - ps->max() + 1);
insert(memSize(), newmax - ps->max());
ps->setMem (newmax + 1, ps->mem());
ps->set_size( sz );
}
else
{
ensureMem(newmax + 1);
SVector newps(newmax + 1, &last() + 1);
int sz = ps->size();
insert(memSize(), newmax + 1);
newps = svec;
if (ps != list.first())
{
SVector* prev = ps->prev();
int prevsz = prev->size();
prev->setMem (prev->max()
+ ps->max() + 2, prev->mem());
prev->set_size(prevsz);
possiblyUnusedMem += ps->max();
}
list.remove(ps);
list.append(ps);
ps->setMem(newmax + 1, newps.mem());
ps->set_size(sz);
}
}
}
void SVSet::add2(SVector &svec, int n, const int idx[], const Real val[])
{
xtend(svec, svec.size() + n);
svec.add(n, idx, val);
}
void SVSet::add2(SVector &svec, int idx, Real val)
{
xtend(svec, svec.size() + 1);
svec.add(idx, val);
}