// =============================================================================
// ----- RepName Class ---------------------------------------------------------
// RepName CONSTRUCTOR
File::RepName::RepName(FileName const& fname, int debugIn) :
extChar_('.')
{
if (debugIn > 1)
mprintf("\tREMDTRAJ: FileName=[%s]\n", fname.full());
if ( fname.Ext().empty() ) {
mprinterr("Error: Traj %s has no numerical extension, required for automatic\n"
"Error: detection of replica trajectories. Expected filename format is\n"
"Error: <Prefix>.<#> (with optional compression extension), examples:\n"
"Error: Rep.traj.nc.000, remd.x.01.gz etc.\n", fname.base());
return;
}
// Split off everything before replica extension
size_t found = fname.Full().rfind( fname.Ext() );
Prefix_.assign( fname.Full().substr(0, found) );
ReplicaExt_.assign( fname.Ext() ); // This should be the numeric extension
// Remove leading '.'
if (ReplicaExt_[0] == '.') ReplicaExt_.erase(0,1);
CompressExt_.assign( fname.Compress() );
if (debugIn > 1) {
mprintf("\tREMDTRAJ: Prefix=[%s], #Ext=[%s], CompressExt=[%s]\n",
Prefix_.c_str(), ReplicaExt_.c_str(), CompressExt_.c_str());
}
// CHARMM replica numbers are format <name>_<num>
if ( !validInteger(ReplicaExt_) ) {
size_t uscore = fname.Full().rfind('_');
if (uscore != std::string::npos) {
Prefix_.assign( fname.Full().substr(0, uscore) );
ReplicaExt_.assign( fname.Full().substr(uscore+1) );
extChar_ = '_';
if (debugIn > 0)
mprintf("\tREMDTRAJ: CHARMM style replica names detected, prefix='%s' ext='%s'\n",
Prefix_.c_str(), ReplicaExt_.c_str());
}
}
// Check that the numerical extension is valid.
if ( !validInteger(ReplicaExt_) ) {
mprinterr("Error: Replica extension [%s] is not an integer.\n", ReplicaExt_.c_str());
Prefix_.clear(); // Empty Prefix_ indicates error.
return;
}
ExtWidth_ = (int)ReplicaExt_.size();
if (debugIn > 1)
mprintf("\tREMDTRAJ: Numerical Extension width=%i\n", ExtWidth_);
// Store lowest replica number
lowestRepnum_ = convertToInteger( ReplicaExt_ );
// TODO: Do not allow negative replica numbers?
if (debugIn > 1)
mprintf("\tREMDTRAJ: index of first replica = %i\n", lowestRepnum_);
}
int DataIO_XVG::ReadData(FileName const& fname,
DataSetList& datasetlist, std::string const& dsname)
{
std::vector<std::string> Legends;
BufferedLine infile;
if (infile.OpenFileRead( fname )) return 1;
const char* ptr = infile.Line();
if (ptr == 0) return 1;
// Skip any comments
while (ptr != 0 && ptr[0] == '#')
ptr = infile.Line();
// Try to get set legends
while (ptr != 0 && ptr[0] == '@') {
ArgList line(ptr, " \t");
if (line.Nargs() > 3 && line[1][0] == 's') {
std::string legend = line.GetStringKey("legend");
if (!legend.empty()) {
// Spaces will cause issues with data set selection.
for (std::string::iterator s = legend.begin(); s != legend.end(); ++s)
if (*s == ' ') *s = '_';
Legends.push_back( legend );
}
}
ptr = infile.Line();
}
if (Legends.empty()) {
mprinterr("Error: No set legends found in XVG file.\n");
return 1;
}
if (ptr == 0) {
mprinterr("Error: No data in XVG file.\n");
return 1;
}
// Create 1 data set for each legend
DataSetList::DataListType inputSets;
for (unsigned int i = 0; i != Legends.size(); i++) {
MetaData md( dsname, i );
md.SetLegend( Legends[i] );
DataSet_double* ds = new DataSet_double();
if (ds == 0) return 1;
ds->SetMeta( md );
inputSets.push_back( ds );
}
mprintf("\t%s has %zu columns of data.\n", fname.base(), inputSets.size());
// Should now be positioned at first line of data. Assume first column is time values.
DataSetList::Darray Xvals;
int expectedCols = (int)inputSets.size() + 1;
while (ptr != 0) {
int ncols = infile.TokenizeLine(" \t");
if (ncols != expectedCols)
mprinterr("Error: Line %i: %i columns != expected # cols %i\n", infile.LineNumber(),
ncols, expectedCols);
else {
Xvals.push_back( atof( infile.NextToken() ) );
for (unsigned int i = 0; i != inputSets.size(); i++)
((DataSet_double*)inputSets[i])->AddElement( atof( infile.NextToken() ) );
}
ptr = infile.Line();
}
infile.CloseFile();
return (datasetlist.AddOrAppendSets( "", Xvals, inputSets ));
}
// Action_Spam::init()
Action::RetType Action_Spam::Init(ArgList& actionArgs, TopologyList* PFL, DataSetList* DSL, DataFileList* DFL, int debugIn)
{
// Always use imaged distances
InitImaging(true);
// This is needed everywhere in this function scope
FileName filename;
// See if we're doing pure water. If so, we don't need a peak file
purewater_ = actionArgs.hasKey("purewater");
if (purewater_) {
// We still need the cutoff
double cut = actionArgs.getKeyDouble("cut", 12.0);
cut2_ = cut * cut;
doublecut_ = 2 * cut;
onecut2_ = 1 / cut2_;
// See if we write to a data file
datafile_ = actionArgs.GetStringKey("out");
// Generate the data set name, and hold onto the master data set list
std::string ds_name = actionArgs.GetStringKey("name");
if (ds_name.empty())
ds_name = myDSL_.GenerateDefaultName("SPAM");
// We only have one data set averaging over every water. Add it here
myDSL_.AddSet(DataSet::DOUBLE, ds_name, NULL);
solvname_ = actionArgs.GetStringKey("solv");
if (solvname_.empty())
solvname_ = std::string("WAT");
}else {
// Get the file name with the peaks defined in it
filename.SetFileName( actionArgs.GetStringNext() );
if (filename.empty() || !File::Exists(filename)) {
mprinterr("Spam: Error: Peak file [%s] does not exist!\n", filename.full());
return Action::ERR;
}
// Get the remaining optional arguments
solvname_ = actionArgs.GetStringKey("solv");
if (solvname_.empty())
solvname_ = std::string("WAT");
reorder_ = actionArgs.hasKey("reorder");
bulk_ = actionArgs.getKeyDouble("bulk", 0.0);
double cut = actionArgs.getKeyDouble("cut", 12.0);
cut2_ = cut * cut;
doublecut_ = 2 * cut;
onecut2_ = 1 / cut2_;
std::string infoname = actionArgs.GetStringKey("info");
if (infoname.empty())
infoname = std::string("spam.info");
infofile_ = DFL->AddCpptrajFile(infoname, "SPAM info");
if (infofile_ == 0) return Action::ERR;
// The default maskstr is the Oxygen atom of the solvent
summaryfile_ = actionArgs.GetStringKey("summary");
// Divide site size by 2 to make it half the edge length (or radius)
site_size_ = actionArgs.getKeyDouble("site_size", 2.5) / 2.0;
sphere_ = actionArgs.hasKey("sphere");
// If it's a sphere, square the radius to compare with
if (sphere_)
site_size_ *= site_size_;
datafile_ = actionArgs.GetStringKey("out");
std::string ds_name = actionArgs.GetStringKey("name");
if (ds_name.empty())
ds_name = myDSL_.GenerateDefaultName("SPAM");
// Parse through the peaks file and extract the peaks
CpptrajFile peakfile;
if (peakfile.OpenRead(filename)) {
mprinterr("SPAM: Error: Could not open %s for reading!\n", filename.full());
return Action::ERR;
}
std::string line = peakfile.GetLine();
int npeaks = 0;
while (!line.empty()) {
if (sscanf(line.c_str(), "%d", &npeaks) != 1) {
line = peakfile.GetLine();
continue;
}
line = peakfile.GetLine();
break;
}
while (!line.empty()) {
double x, y, z, dens;
if (sscanf(line.c_str(), "C %lg %lg %lg %lg", &x, &y, &z, &dens) != 4) {
line = peakfile.GetLine();
continue;
}
line = peakfile.GetLine();
peaks_.push_back(Vec3(x, y, z));
}
peakfile.CloseFile();
// Check that our initial number of peaks matches our parsed peaks. Warn
// otherwise
if (npeaks != (int)peaks_.size())
mprinterr("SPAM: Warning: %s claims to have %d peaks, but really has %d!\n",
filename.full(), npeaks, peaks_.size());
// Now add all of the data sets
MetaData md(ds_name);
for (int i = 0; i < (int)peaks_.size(); i++) {
md.SetAspect( integerToString(i+1) ); // TODO: Should this be Idx?
if (myDSL_.AddSet(DataSet::DOUBLE, md) == 0) return Action::ERR;
// Add a new list of integers to keep track of omitted frames
std::vector<int> vec;
peakFrameData_.push_back(vec);
}
}
// Print info now
if (purewater_) {
mprintf("SPAM: Calculating bulk value for pure solvent\n");
if (!datafile_.empty())
mprintf("SPAM: Printing solvent energies to %s\n", datafile_.c_str());
mprintf("SPAM: Using a %.2f Angstrom non-bonded cutoff with shifted EEL.\n",
sqrt(cut2_));
if (reorder_)
mprintf("SPAM: Warning: Re-ordering makes no sense for pure solvent.\n");
if (!summaryfile_.empty())
mprintf("SPAM: Printing solvent SPAM summary to %s\n", summaryfile_.c_str());
}else {
mprintf("SPAM: Solvent [%s] density peaks taken from %s.\n",
solvname_.c_str(), filename.base());
mprintf("SPAM: %d density peaks will be analyzed from %s.\n",
peaks_.size(), filename.base());
mprintf("SPAM: Occupation information printed to %s.\n", infofile_->Filename().full());
mprintf("SPAM: Sites are ");
if (sphere_)
mprintf("spheres with diameter %.3lf\n", site_size_);
else
mprintf("boxes with edge length %.3lf\n", site_size_);
if (reorder_) {
mprintf("SPAM: Re-ordering trajectory so each site always has ");
mprintf("the same water molecule.\n");
}
if (summaryfile_.empty() && datafile_.empty()) {
if (!reorder_) {
mprinterr("SPAM: Error: Not re-ordering trajectory or calculating energies. ");
mprinterr("Nothing to do!\n");
return Action::ERR;
}
mprintf("SPAM: Not calculating any SPAM energies\n");
}else {
mprintf("SPAM: Using a non-bonded cutoff of %.2lf Ang. with a EEL shifting function.\n",
sqrt(cut2_));
mprintf("SPAM: Bulk solvent SPAM energy taken as %.3lf kcal/mol\n", bulk_);
}
}
mprintf("#Citation: Cui, G.; Swails, J.M.; Manas, E.S.; \"SPAM: A Simple Approach\n"
"# for Profiling Bound Water Molecules\"\n"
"# J. Chem. Theory Comput., 2013, 9 (12), pp 5539–5549.\n");
return Action::OK;
}
