void testMetadata(MSDataCache& cache)
{
if (os_) *os_ << "testMetadata()\n";
if (os_) *os_ << "spectrumCount: " << cache.size() << endl;
unit_assert(cache.size() == 5);
unit_assert(cache[0].index == 0);
unit_assert(cache[0].id == "scan=19");
unit_assert(cache[0].scanNumber == 19); // TODO: change to nativeID
unit_assert(cache[0].massAnalyzerType == MS_QIT);
unit_assert(cache[0].msLevel == 1);
unit_assert_equal(cache[0].retentionTime, 353.43, epsilon_);
unit_assert_equal(cache[0].mzLow, 400.39, epsilon_);
unit_assert_equal(cache[0].mzHigh, 1795.56, epsilon_);
unit_assert(cache[0].precursors.empty());
unit_assert(cache[1].index == 1);
unit_assert(cache[1].id == "scan=20");
unit_assert(cache[1].scanNumber == 20); // TODO: change to nativeID
unit_assert(cache[1].massAnalyzerType == MS_QIT);
unit_assert(cache[1].msLevel == 2);
unit_assert_equal(cache[1].retentionTime, 359.43, epsilon_);
unit_assert_equal(cache[1].mzLow, 320.39, epsilon_);
unit_assert_equal(cache[1].mzHigh, 1003.56, epsilon_);
unit_assert(cache[1].precursors.size() == 1);
unit_assert(cache[1].precursors[0].index == 0);
unit_assert_equal(cache[1].precursors[0].mz, 445.34, epsilon_);
unit_assert_equal(cache[1].precursors[0].intensity, 120053, epsilon_);
unit_assert(cache[1].precursors[0].charge == 2);
if (os_) *os_ << endl;
}
void printCache(ostream& os, const MSDataCache& cache)
{
os << "cached binary data:\n";
for (vector<SpectrumInfo>::const_iterator it=cache.begin(); it!=cache.end(); ++it)
{
os << it->index << " "
<< it->data.size() << "/"
<< it->data.capacity() << endl;
}
os << endl;
}
void testAutomaticUpdate()
{
if (os_) *os_ << "testAutomaticUpdate()\n";
vector<MZIntensityPair> pairs(100);
SpectrumListSimplePtr sl(new SpectrumListSimple);
for (size_t i=0; i<10; i++)
{
SpectrumPtr spectrum(new Spectrum);
spectrum->setMZIntensityPairs(pairs, MS_number_of_detector_counts);
spectrum->index = i;
spectrum->id = "scan=" + lexical_cast<string>(i);
sl->spectra.push_back(spectrum);
}
MSData msd;
msd.run.spectrumListPtr = sl;
MSDataCache cache;
cache.open(msd);
unit_assert(cache.size() == sl->size());
for (size_t i=0; i<cache.size(); i++)
unit_assert(cache[i].index == (size_t)-1);
const SpectrumInfo& info5 = cache.spectrumInfo(5, true);
unit_assert(cache[5].data.size() == 100);
cache.spectrumInfo(7); // getBinaryData==false -> doesn't change cached binary data
unit_assert(cache[5].data.size() == 100);
unit_assert(cache[7].data.size() == 0);
const SpectrumInfo& info7 = cache.spectrumInfo(7, true);
if (os_)
{
for (size_t i=0; i<cache.size(); i++)
*os_ << i << " " << cache[i].index << " " << cache[i].id << " "
<< cache[i].data.size() << endl;
}
unit_assert(info7.data.size() == 100);
unit_assert(info5.data.size() == 0);
unit_assert(info5.index==5 && info5.id=="scan=5");
unit_assert(cache[5].index==5 && cache[5].id=="scan=5");
unit_assert(info7.index==7 && info7.id=="scan=7");
unit_assert(cache[7].index==7 && cache[7].id=="scan=7");
for (size_t i=0; i<cache.size(); i++)
if (i!=5 && i!=7)
unit_assert(cache[i].index == (size_t)-1);
}
void testDefault()
{
MSData tiny;
examples::initializeTiny(tiny);
MSDataCache cache;
cache.open(tiny);
cache.update(tiny, *tiny.run.spectrumListPtr->spectrum(0));
unit_assert(!cache[0].data.empty());
unit_assert(cache[1].data.empty());
cache.update(tiny, *tiny.run.spectrumListPtr->spectrum(1));
unit_assert(cache[0].data.empty());
unit_assert(!cache[1].data.empty());
testMetadata(cache);
}
void FeatureDetectorSimple::Impl::detect(const MSData& msd, FeatureField& detected) const
{
// initialize buffer maps
map<FeatureKey, FeatureID> grandparentBuffer;
map<FeatureKey, FeatureID> parentBuffer;
map<FeatureID, FeaturePtr> featureMap;
size_t featureCount = 0;
MSDataCache cache;
cache.open(msd);
for(size_t spectrum_index = 0; spectrum_index < cache.size(); spectrum_index ++)
{
map<FeatureKey, FeatureID> buffer;
const SpectrumInfo info = cache.spectrumInfo(spectrum_index, true); //getBinaryData ?
// if info.PrecursorInfo ?
// call peak family detector on each scan
vector<MZIntensityPair> mzIntensityPairs = info.data;
vector<PeakFamily> result;
if (info.massAnalyzerType != MS_FT_ICR && info.massAnalyzerType != MS_orbitrap)
{
cerr << "Skipping non-FT scan number " << info.scanNumber << endl;
continue;
}
_pfd->detect(mzIntensityPairs,result);
// iterate thru peak families
vector<PeakFamily>::iterator result_it = result.begin();
double rt = info.retentionTime;
int id = info.scanNumber;
for(; result_it != result.end(); ++result_it)
{
// set ID attribute
for_each(result_it->peaks.begin(), result_it->peaks.end(), SetID(id));
// set RT attribute
for_each(result_it->peaks.begin(), result_it->peaks.end(), SetRT(rt));
// make keys for search
FeatureKey featureKey(floor(result_it->mzMonoisotopic*100)/100, result_it->charge, result_it->peaks.size()); // no digits past 10e-3
map<FeatureKey,FeatureID>::iterator grandparentLocation = grandparentBuffer.find(featureKey);
map<FeatureKey,FeatureID>::iterator parentLocation = parentBuffer.find(featureKey);
if ( parentLocation == parentBuffer.end() && grandparentLocation == grandparentBuffer.end())
{
// feature doesn't exist, make new feature
FeatureID featureID = boost::lexical_cast<string>(++featureCount);
FeaturePtr feature(makeFeature(*result_it,featureID));
featureMap.insert(pair<FeatureID, FeaturePtr>(featureID, feature)); buffer.insert(pair<FeatureKey, FeatureID>(featureKey, featureID));
}
else
{
// if parent location update from parent
FeatureID foundID;
if ( parentLocation != parentBuffer.end() ) foundID = parentLocation->second;
else foundID = grandparentLocation->second;
// update existing feature
updateFeature(*result_it, featureMap[foundID]);
buffer.insert(pair<FeatureKey, FeatureID>(featureKey,foundID));
}
}
// old buffer map := new buffer map
// bufferMap.swap(updatedBufferMap);
grandparentBuffer.swap(parentBuffer);
parentBuffer.swap(buffer);
}
// write output vector<Feature>
map<FeatureID, FeaturePtr>::iterator the_it = featureMap.begin();
for(; the_it != featureMap.end(); ++the_it)
{
if(the_it->second->peakels.front()->peaks.size()>1) // for now only write features lasting > 1 scan
{
the_it->second->calculateMetadata();
detected.insert(the_it->second);
}
}
return;
}