// Action_NativeContacts::WriteContacts()
void Action_NativeContacts::WriteContacts(contactListType& ContactsIn, bool isNative) {
if (ContactsIn.empty()) return;
// Map of residue pairs to total contact values.
typedef std::map<Cpair, resContact> resContactMap;
resContactMap ResContacts;
std::pair<resContactMap::iterator, bool> ret;
// Normalize native contacts. Place them into an array where they will
// be sorted. Sum up total contact over residue pairs.
std::vector<contactType> sortedList;
for (contactListType::iterator it = ContactsIn.begin();
it != ContactsIn.end(); ++it)
{
it->second.Finalize();
sortedList.push_back( it->second );
ret = ResContacts.insert( Rpair(Cpair(it->second.Res1(),it->second.Res2()),
resContact(it->second.Nframes(), it->second.DataPtr())) );
if (!ret.second) // residue pair exists, update it.
ret.first->second.Increment( it->second.Nframes(), it->second.DataPtr() );
}
std::sort( sortedList.begin(), sortedList.end() );
// Place residue pairs into an array to be sorted.
std::vector<Rpair> ResList;
for (resContactMap::const_iterator it = ResContacts.begin(); it != ResContacts.end(); ++it)
{
ResList.push_back( *it );
if (Rseries_ != NO_RESSERIES) {
const char* resDsAspect;
std::string lprefix = "";
if (isNative)
resDsAspect = "NCRES";
else {
resDsAspect = "NNRES";
lprefix = "nn_";
}
// Ensure r1 < r2 so we can calculate a unique index for residue pairs
// that can match between native/non-native contacts
int r1, r2;
if (it->first.second < it->first.first) {
r1 = it->first.second;
r2 = it->first.first;
} else {
r1 = it->first.first;
r2 = it->first.second;
}
int ridx = (r2 * CurrentParm_->Nres()) + r1;
std::string legend(lprefix +
CurrentParm_->TruncResNameNum(r1) + "_" +
CurrentParm_->TruncResNameNum(r2));
MetaData md(numnative_->Meta().Name(), resDsAspect, ridx);
md.SetLegend( legend );
DataSet_integer* ds = (DataSet_integer*)masterDSL_->AddSet(DataSet::INTEGER, md);
if (ds != 0) {
ds->Allocate(DataSet::SizeArray(1, nframes_));
if (seriesRout_ != 0) seriesRout_->AddDataSet( ds );
// All series will be the same size thanks to UpdateSeries()
for (unsigned int f = 0; f != nframes_; f++) {
int total_present = 0;
for (DSarray::const_iterator set = it->second.Sets().begin();
set != it->second.Sets().end(); ++set)
total_present += (*(*set))[f];
if (Rseries_ == RES_PRESENT && total_present > 0) total_present = 1;
ds->AddElement( total_present );
}
}
}
}
std::sort( ResList.begin(), ResList.end(), res_cmp() );
// Print out total fraction frames for residue pairs.
const char* ctitle;
if (isNative)
ctitle = "Contacts";
else
ctitle = "nnContacts";
rfile_->Printf("%-8s %8s %10s %10s\n", "#Res1", "#Res2", "TotalFrac", ctitle);
//for (resContactMap::const_iterator it = ResContacts.begin(); it != ResContacts.end(); ++it)
for (std::vector<Rpair>::const_iterator it = ResList.begin();
it != ResList.end(); ++it)
rfile_->Printf("%-8i %8i %10g %10i\n", it->first.first+1, it->first.second+1,
(double)it->second.Nframes()/(double)nframes_,
it->second.Ncontacts());
// Print out sorted atom contacts.
cfile_->Printf("%-8s %20s %8s %8s %8s %8s\n", "#", "Contact", "Nframes", "Frac.", "Avg", "Stdev");
unsigned int num = 1;
for (std::vector<contactType>::const_iterator NC = sortedList.begin();
NC != sortedList.end(); ++NC, ++num)
{
double fracPresent = (double)NC->Nframes() / (double)nframes_;
cfile_->Printf("%8u %20s %8i %8.3g %8.3g %8.3g\n", num, NC->id(),
NC->Nframes(), fracPresent, NC->Avg(), NC->Stdev());
}
}
// Analysis_RemLog::Setup()
Analysis::RetType Analysis_RemLog::Setup(ArgList& analyzeArgs, DataSetList* datasetlist, DataFileList* DFLin, int debugIn)
{
debug_ = debugIn;
// Get remlog dataset
std::string remlogName = analyzeArgs.GetStringNext();
if (remlogName.empty()) {
mprinterr("Error: no remlog data set or file name specified.\n");
return Analysis::ERR;
}
// Check if data set exists
remlog_ = (DataSet_RemLog*)datasetlist->FindSetOfType( remlogName, DataSet::REMLOG );
if (remlog_ == 0) {
mprinterr("Error: remlog data with name %s not found.\n", remlogName.c_str());
return Analysis::ERR;
}
if (remlog_->Size() < 1 || remlog_->NumExchange() < 1) {
mprinterr("Error: remlog data set appears to be empty.\n");
return Analysis::ERR;
}
acceptout_ = DFLin->AddCpptrajFile( analyzeArgs.GetStringKey("acceptout"), "replica acceptance",
DataFileList::TEXT, true );
if (acceptout_ == 0) return Analysis::ERR;
lifetimes_ = DFLin->AddCpptrajFile( analyzeArgs.GetStringKey("lifetime"), "remlog lifetimes" );
calculateLifetimes_ = (lifetimes_ != 0);
calculateStats_ = analyzeArgs.hasKey("stats");
if (calculateStats_) {
statsout_ = DFLin->AddCpptrajFile( analyzeArgs.GetStringKey("statsout"), "remlog stats",
DataFileList::TEXT, true );
reptime_ = DFLin->AddCpptrajFile( analyzeArgs.GetStringKey("reptime"), "replica times",
DataFileList::TEXT, true );
if (statsout_ == 0 || reptime_ == 0) return Analysis::ERR;
}
calcRepFracSlope_ = analyzeArgs.getKeyInt("reptimeslope", 0);
std::string rfs_name = analyzeArgs.GetStringKey("reptimeslopeout");
if (!calculateStats_) {
calcRepFracSlope_ = 0;
rfs_name.clear();
}
if ( (calcRepFracSlope_ > 0) != (!rfs_name.empty()) ) {
mprinterr("Error: Both reptimeslope and reptimeslopeout must be specified.\n");
return Analysis::ERR;
}
repFracSlope_ = DFLin->AddCpptrajFile( rfs_name, "replica fraction slope" );
printIndividualTrips_ = analyzeArgs.hasKey("printtrips");
// Get mode
if (analyzeArgs.hasKey("crdidx"))
mode_ = CRDIDX;
else if (analyzeArgs.hasKey("repidx"))
mode_ = REPIDX;
else
mode_ = NONE;
const char* def_name = 0;
const char* yaxis = 0;
if (mode_ == CRDIDX) {
def_name = "repidx";
yaxis = "ylabel CrdIdx";
} else if (mode_ == REPIDX) {
def_name = "crdidx";
yaxis = "ylabel RepIdx";
}
// Set up an output set for each replica
DataFile* dfout = 0;
if (mode_ != NONE) {
// Get output filename
std::string outname = analyzeArgs.GetStringKey("out");
if (!outname.empty()) {
dfout = DFLin->AddDataFile( outname, analyzeArgs );
if (dfout == 0 ) return Analysis::ERR;
if (yaxis != 0 ) dfout->ProcessArgs(yaxis);
}
std::string dsname = analyzeArgs.GetStringKey("name");
if (dsname.empty())
dsname = datasetlist->GenerateDefaultName(def_name);
MetaData md(dsname);
for (int i = 0; i < (int)remlog_->Size(); i++) {
md.SetIdx(i+1);
DataSet_integer* ds = (DataSet_integer*)datasetlist->AddSet(DataSet::INTEGER, md);
if (ds == 0) return Analysis::ERR;
outputDsets_.push_back( (DataSet*)ds );
if (dfout != 0) dfout->AddDataSet( (DataSet*)ds );
ds->Resize( remlog_->NumExchange() );
}
}
mprintf(" REMLOG: %s, %i replicas, %i exchanges\n", remlog_->legend(),
remlog_->Size(), remlog_->NumExchange());
if (mode_ == CRDIDX)
mprintf("\tGetting coordinate index vs exchange.\n");
else if (mode_ == REPIDX)
mprintf("\tGetting replica index vs exchange.\n");
if (mode_ != NONE && dfout != 0)
mprintf("\tOutput is to %s\n", dfout->DataFilename().base());
if (calculateStats_) {
mprintf("\tGetting replica exchange stats, output to %s\n", statsout_->Filename().full());
if (printIndividualTrips_)
mprintf("\tIndividual round trips will be printed.\n");
mprintf("\tWriting time spent at each replica to %s\n", reptime_->Filename().full());
}
if (calculateLifetimes_)
mprintf("\tThe lifetime of each crd at each replica will be calculated.\n");
if (acceptout_ != 0)
mprintf("\tOverall exchange acceptance % will be written to %s\n",
acceptout_->Filename().full());
return Analysis::OK;
}
// Action_ClusterDihedral::Print()
void Action_ClusterDihedral::Print() {
// Setup output file
mprintf("\tPrinting Dihedral Clustering Results.\n");
// Print bin information
outfile_->Printf("DIHEDRAL CLUSTER RESULTS");
if (mask_.MaskStringSet())
outfile_->Printf(" for %s", mask_.MaskString());
outfile_->Printf("\n");
long int num = 0;
for (std::vector<DCmask>::const_iterator dih = DCmasks_.begin();
dih != DCmasks_.end(); ++dih)
{
outfile_->Printf(" %6li ", num++);
outfile_->Printf("%-s(%i)", (*dcparm_)[ dih->A1() ].c_str(), dih->A1() + 1);
outfile_->Printf("%-s(%i)", (*dcparm_)[ dih->A2() ].c_str(), dih->A2() + 1);
outfile_->Printf("%-s(%i)", (*dcparm_)[ dih->A3() ].c_str(), dih->A3() + 1);
outfile_->Printf("%-s(%i)", (*dcparm_)[ dih->A4() ].c_str(), dih->A4() + 1);
outfile_->Printf(" [Bins=%i]\n", dih->Bins());
}
outfile_->Printf("%zu clusters.\n", dcarray_.size());
// Sort array by count
std::sort( dcarray_.begin(), dcarray_.end() );
// Allocate space for storing cluster #s for each frame
std::vector<long int> framecluster( lastframe_ + 1 );
// Print sorted cluster array
if (CUT_ > 0)
outfile_->Printf("Only printing clusters with pop > %i\n",CUT_);
num = 0;
for (std::vector<DCnode>::const_iterator DC = dcarray_.begin();
DC != dcarray_.end(); ++DC)
{
if ( DC->Count() > CUT_ ) {
//mprintf("DEBUG: Cluster %li has %i frames.\n", num, DC->NumFrames());
outfile_->Printf("Cluster %10li %10li [ ", num+1, DC->Count());
for (DCnode::bin_it binid = DC->binbegin(); binid != DC->binend(); ++binid)
outfile_->Printf("%3i ", *binid);
outfile_->Printf(" ]\n");
for (DCnode::frame_it frame = DC->framebegin(); frame != DC->frameend(); ++frame)
{
outfile_->Printf("%i ", *frame + 1);
// store which cluster each frame belongs to. Not neccesary if user
// didn't specify this option, but avoids a second loop if they did.
framecluster[ *frame ] = num;
}
outfile_->Printf("\n");
}
++num;
}
// Place reordered cluster nums in CVT
if (CVT_ != 0) {
DataSet_integer* iCVT = (DataSet_integer*)CVT_;
iCVT->Resize( framecluster.size() );
num = 0;
for (std::vector<long int>::const_iterator cnum = framecluster.begin();
cnum != framecluster.end(); ++cnum)
(*iCVT)[ num++ ] = (int)*cnum + 1;
}
// Print cluster for each frame
if (framefile_ != 0) {
mprintf("\tPrinting cluster number for each frame.\n");
num = 1;
for (std::vector<long int>::const_iterator cnum = framecluster.begin();
cnum != framecluster.end(); ++cnum)
{
// Frame, cluster num, cluster count
framefile_->Printf("%10li %10i %10li ", num++, *cnum + 1, dcarray_[*cnum].Count());
// Print binID
for (DCnode::bin_it binid = dcarray_[*cnum].binbegin();
binid != dcarray_[*cnum].binend(); ++binid)
framefile_->Printf("%03i", *binid);
framefile_->Printf("\n");
}
}
// Print cluster information file
if (infofile_!=0) {
mprintf("\tPrinting cluster information.\n");
infofile_->Printf("%zu\n", DCmasks_.size());
for (std::vector<DCmask>::const_iterator dih = DCmasks_.begin();
dih != DCmasks_.end(); ++dih)
{
infofile_->Printf("%10i %10i %10i %10i %10i %8.3f\n", dih->A1()+1, dih->A2()+1,
dih->A3()+1, dih->A4()+1, dih->Bins(), dih->Min());
}
infofile_->Printf("%zu\n", dcarray_.size());
num = 1;
for (std::vector<DCnode>::const_iterator DC = dcarray_.begin();
DC != dcarray_.end(); ++DC)
{
infofile_->Printf("%10li %10li ", num++, DC->Count());
for (DCnode::bin_it binid = DC->binbegin(); binid != DC->binend(); ++binid)
infofile_->Printf(" %3i", *binid);
infofile_->Printf("\n");
}
}
}
