src/gromacs/trajectoryanalysis/modules/distance.cpp

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2010,2011,2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*! \internal \file
  36  * \brief
  37  * Implements gmx::analysismodules::Distance.
  38  *
  39  * \author Teemu Murtola <teemu.murtola@gmail.com>
  40  * \ingroup module_trajectoryanalysis
  41  */
  42 #include "gmxpre.h"
  43
  44 #include "distance.h"
  45
  46 #include <string>
  47
  48 #include "gromacs/analysisdata/analysisdata.h"
  49 #include "gromacs/analysisdata/modules/average.h"
  50 #include "gromacs/analysisdata/modules/histogram.h"
  51 #include "gromacs/analysisdata/modules/plot.h"
  52 #include "gromacs/fileio/trx.h"
  53 #include "gromacs/math/vec.h"
  54 #include "gromacs/options/basicoptions.h"
  55 #include "gromacs/options/filenameoption.h"
  56 #include "gromacs/options/options.h"
  57 #include "gromacs/pbcutil/pbc.h"
  58 #include "gromacs/selection/selection.h"
  59 #include "gromacs/selection/selectionoption.h"
  60 #include "gromacs/trajectoryanalysis/analysissettings.h"
  61 #include "gromacs/utility/arrayref.h"
  62 #include "gromacs/utility/exceptions.h"
  63 #include "gromacs/utility/stringutil.h"
  64
  65 namespace gmx
  66 {
  67
  68 namespace analysismodules
  69 {
  70
  71 namespace
  72 {
  73
  74 class Distance : public TrajectoryAnalysisModule
  75 {
  76     public:
  77         Distance();
  78
  79         virtual void initOptions(Options                    *options,
  80                                  TrajectoryAnalysisSettings *settings);
  81         virtual void initAnalysis(const TrajectoryAnalysisSettings &settings,
  82                                   const TopologyInformation        &top);
  83
  84         virtual void analyzeFrame(int frnr, const t_trxframe &fr, t_pbc *pbc,
  85                                   TrajectoryAnalysisModuleData *pdata);
  86
  87         virtual void finishAnalysis(int nframes);
  88         virtual void writeOutput();
  89
  90     private:
  91         SelectionList                            sel_;
  92         std::string                              fnAverage_;
  93         std::string                              fnAll_;
  94         std::string                              fnXYZ_;
  95         std::string                              fnHistogram_;
  96         std::string                              fnAllStats_;
  97         double                                   meanLength_;
  98         double                                   lengthDev_;
  99         double                                   binWidth_;
 100
 101         AnalysisData                             distances_;
 102         AnalysisData                             xyz_;
 103         AnalysisDataAverageModulePointer         summaryStatsModule_;
 104         AnalysisDataAverageModulePointer         allStatsModule_;
 105         AnalysisDataFrameAverageModulePointer    averageModule_;
 106         AnalysisDataSimpleHistogramModulePointer histogramModule_;
 107
 108         // Copy and assign disallowed by base.
 109 };
 110
 111 Distance::Distance()
 112     : TrajectoryAnalysisModule(DistanceInfo::name, DistanceInfo::shortDescription),
 113       meanLength_(0.1), lengthDev_(1.0), binWidth_(0.001)
 114 {
 115     summaryStatsModule_.reset(new AnalysisDataAverageModule());
 116     summaryStatsModule_->setAverageDataSets(true);
 117     distances_.addModule(summaryStatsModule_);
 118     allStatsModule_.reset(new AnalysisDataAverageModule());
 119     distances_.addModule(allStatsModule_);
 120     averageModule_.reset(new AnalysisDataFrameAverageModule());
 121     distances_.addModule(averageModule_);
 122     histogramModule_.reset(new AnalysisDataSimpleHistogramModule());
 123     distances_.addModule(histogramModule_);
 124
 125     registerAnalysisDataset(&distances_, "dist");
 126     registerAnalysisDataset(&xyz_, "xyz");
 127     registerBasicDataset(summaryStatsModule_.get(), "stats");
 128     registerBasicDataset(allStatsModule_.get(), "allstats");
 129     registerBasicDataset(averageModule_.get(), "average");
 130     registerBasicDataset(&histogramModule_->averager(), "histogram");
 131 }
 132
 133
 134 void
 135 Distance::initOptions(Options *options, TrajectoryAnalysisSettings * /*settings*/)
 136 {
 137     static const char *const desc[] = {
 138         "[THISMODULE] calculates distances between pairs of positions",
 139         "as a function of time. Each selection specifies an independent set",
 140         "of distances to calculate. Each selection should consist of pairs",
 141         "of positions, and the distances are computed between positions 1-2,",
 142         "3-4, etc.[PAR]",
 143         "[TT]-oav[tt] writes the average distance as a function of time for",
 144         "each selection.",
 145         "[TT]-oall[tt] writes all the individual distances.",
 146         "[TT]-oxyz[tt] does the same, but the x, y, and z components of the",
 147         "distance are written instead of the norm.",
 148         "[TT]-oh[tt] writes a histogram of the distances for each selection.",
 149         "The location of the histogram is set with [TT]-len[tt] and",
 150         "[TT]-tol[tt]. Bin width is set with [TT]-binw[tt].",
 151         "[TT]-oallstat[tt] writes out the average and standard deviation for",
 152         "each individual distance, calculated over the frames."
 153     };
 154
 155     options->setDescription(desc);
 156
 157     options->addOption(FileNameOption("oav").filetype(eftPlot).outputFile()
 158                            .store(&fnAverage_).defaultBasename("distave")
 159                            .description("Average distances as function of time"));
 160     options->addOption(FileNameOption("oall").filetype(eftPlot).outputFile()
 161                            .store(&fnAll_).defaultBasename("dist")
 162                            .description("All distances as function of time"));
 163     options->addOption(FileNameOption("oxyz").filetype(eftPlot).outputFile()
 164                            .store(&fnXYZ_).defaultBasename("distxyz")
 165                            .description("Distance components as function of time"));
 166     options->addOption(FileNameOption("oh").filetype(eftPlot).outputFile()
 167                            .store(&fnHistogram_).defaultBasename("disthist")
 168                            .description("Histogram of the distances"));
 169     options->addOption(FileNameOption("oallstat").filetype(eftPlot).outputFile()
 170                            .store(&fnAllStats_).defaultBasename("diststat")
 171                            .description("Statistics for individual distances"));
 172     options->addOption(SelectionOption("select").storeVector(&sel_)
 173                            .required().dynamicMask().multiValue()
 174                            .description("Position pairs to calculate distances for"));
 175     // TODO: Extend the histogramming implementation to allow automatic
 176     // extension of the histograms to cover the data, removing the need for
 177     // the first two options.
 178     options->addOption(DoubleOption("len").store(&meanLength_)
 179                            .description("Mean distance for histogramming"));
 180     options->addOption(DoubleOption("tol").store(&lengthDev_)
 181                            .description("Width of full distribution as fraction of [TT]-len[tt]"));
 182     options->addOption(DoubleOption("binw").store(&binWidth_)
 183                            .description("Bin width for histogramming"));
 184 }
 185
 186
 187 /*! \brief
 188  * Checks that selections conform to the expectations of the tool.
 189  */
 190 void checkSelections(const SelectionList &sel)
 191 {
 192     for (size_t g = 0; g < sel.size(); ++g)
 193     {
 194         if (sel[g].posCount() % 2 != 0)
 195         {
 196             std::string message = formatString(
 197                         "Selection '%s' does not evaluate into an even number of positions "
 198                         "(there are %d positions)",
 199                         sel[g].name(), sel[g].posCount());
 200             GMX_THROW(InconsistentInputError(message));
 201         }
 202         if (sel[g].isDynamic())
 203         {
 204             for (int i = 0; i < sel[g].posCount(); i += 2)
 205             {
 206                 if (sel[g].position(i).selected() != sel[g].position(i+1).selected())
 207                 {
 208                     std::string message =
 209                         formatString("Dynamic selection %d does not select "
 210                                      "a consistent set of pairs over the frames",
 211                                      static_cast<int>(g + 1));
 212                     GMX_THROW(InconsistentInputError(message));
 213                 }
 214             }
 215         }
 216     }
 217 }
 218
 219
 220 void
 221 Distance::initAnalysis(const TrajectoryAnalysisSettings &settings,
 222                        const TopologyInformation         & /*top*/)
 223 {
 224     checkSelections(sel_);
 225
 226     distances_.setDataSetCount(sel_.size());
 227     xyz_.setDataSetCount(sel_.size());
 228     for (size_t i = 0; i < sel_.size(); ++i)
 229     {
 230         const int distCount = sel_[i].posCount() / 2;
 231         distances_.setColumnCount(i, distCount);
 232         xyz_.setColumnCount(i, distCount * 3);
 233     }
 234     const double histogramMin = (1.0 - lengthDev_) * meanLength_;
 235     const double histogramMax = (1.0 + lengthDev_) * meanLength_;
 236     histogramModule_->init(histogramFromRange(histogramMin, histogramMax)
 237                                .binWidth(binWidth_).includeAll());
 238
 239     if (!fnAverage_.empty())
 240     {
 241         AnalysisDataPlotModulePointer plotm(
 242                 new AnalysisDataPlotModule(settings.plotSettings()));
 243         plotm->setFileName(fnAverage_);
 244         plotm->setTitle("Average distance");
 245         plotm->setXAxisIsTime();
 246         plotm->setYLabel("Distance (nm)");
 247         for (size_t g = 0; g < sel_.size(); ++g)
 248         {
 249             plotm->appendLegend(sel_[g].name());
 250         }
 251         averageModule_->addModule(plotm);
 252     }
 253
 254     if (!fnAll_.empty())
 255     {
 256         AnalysisDataPlotModulePointer plotm(
 257                 new AnalysisDataPlotModule(settings.plotSettings()));
 258         plotm->setFileName(fnAll_);
 259         plotm->setTitle("Distance");
 260         plotm->setXAxisIsTime();
 261         plotm->setYLabel("Distance (nm)");
 262         // TODO: Add legends? (there can be a massive amount of columns)
 263         distances_.addModule(plotm);
 264     }
 265
 266     if (!fnXYZ_.empty())
 267     {
 268         AnalysisDataPlotModulePointer plotm(
 269                 new AnalysisDataPlotModule(settings.plotSettings()));
 270         plotm->setFileName(fnAll_);
 271         plotm->setTitle("Distance");
 272         plotm->setXAxisIsTime();
 273         plotm->setYLabel("Distance (nm)");
 274         // TODO: Add legends? (there can be a massive amount of columns)
 275         xyz_.addModule(plotm);
 276     }
 277
 278     if (!fnHistogram_.empty())
 279     {
 280         AnalysisDataPlotModulePointer plotm(
 281                 new AnalysisDataPlotModule(settings.plotSettings()));
 282         plotm->setFileName(fnHistogram_);
 283         plotm->setTitle("Distance histogram");
 284         plotm->setXLabel("Distance (nm)");
 285         plotm->setYLabel("Probability");
 286         for (size_t g = 0; g < sel_.size(); ++g)
 287         {
 288             plotm->appendLegend(sel_[g].name());
 289         }
 290         histogramModule_->averager().addModule(plotm);
 291     }
 292
 293     if (!fnAllStats_.empty())
 294     {
 295         AnalysisDataPlotModulePointer plotm(
 296                 new AnalysisDataPlotModule(settings.plotSettings()));
 297         plotm->setFileName(fnAllStats_);
 298         plotm->setErrorsAsSeparateColumn(true);
 299         plotm->setTitle("Statistics for individual distances");
 300         plotm->setXLabel("Distance index");
 301         plotm->setYLabel("Average/standard deviation (nm)");
 302         for (size_t g = 0; g < sel_.size(); ++g)
 303         {
 304             plotm->appendLegend(std::string(sel_[g].name()) + " avg");
 305             plotm->appendLegend(std::string(sel_[g].name()) + " std.dev.");
 306         }
 307         // TODO: Consider whether this output format is the best possible.
 308         allStatsModule_->addModule(plotm);
 309     }
 310 }
 311
 312
 313 void
 314 Distance::analyzeFrame(int frnr, const t_trxframe &fr, t_pbc *pbc,
 315                        TrajectoryAnalysisModuleData *pdata)
 316 {
 317     AnalysisDataHandle   distHandle = pdata->dataHandle(distances_);
 318     AnalysisDataHandle   xyzHandle  = pdata->dataHandle(xyz_);
 319     const SelectionList &sel        = pdata->parallelSelections(sel_);
 320
 321     checkSelections(sel);
 322
 323     distHandle.startFrame(frnr, fr.time);
 324     xyzHandle.startFrame(frnr, fr.time);
 325     for (size_t g = 0; g < sel.size(); ++g)
 326     {
 327         distHandle.selectDataSet(g);
 328         xyzHandle.selectDataSet(g);
 329         for (int i = 0, n = 0; i < sel[g].posCount(); i += 2, ++n)
 330         {
 331             const SelectionPosition &p1 = sel[g].position(i);
 332             const SelectionPosition &p2 = sel[g].position(i+1);
 333             rvec                     dx;
 334             if (pbc != NULL)
 335             {
 336                 pbc_dx(pbc, p2.x(), p1.x(), dx);
 337             }
 338             else
 339             {
 340                 rvec_sub(p2.x(), p1.x(), dx);
 341             }
 342             real dist     = norm(dx);
 343             bool bPresent = p1.selected() && p2.selected();
 344             distHandle.setPoint(n, dist, bPresent);
 345             xyzHandle.setPoints(n*3, 3, dx);
 346         }
 347     }
 348     distHandle.finishFrame();
 349     xyzHandle.finishFrame();
 350 }
 351
 352
 353 void
 354 Distance::finishAnalysis(int /*nframes*/)
 355 {
 356     AbstractAverageHistogram &averageHistogram = histogramModule_->averager();
 357     averageHistogram.normalizeProbability();
 358     averageHistogram.done();
 359 }
 360
 361
 362 void
 363 Distance::writeOutput()
 364 {
 365     SelectionList::const_iterator sel;
 366     int                           index;
 367     for (sel = sel_.begin(), index = 0; sel != sel_.end(); ++sel, ++index)
 368     {
 369         printf("%s:\n", sel->name());
 370         printf("  Number of samples:  %d\n",
 371                summaryStatsModule_->sampleCount(index, 0));
 372         printf("  Average distance:   %-8.5f nm\n",
 373                summaryStatsModule_->average(index, 0));
 374         printf("  Standard deviation: %-8.5f nm\n",
 375                summaryStatsModule_->standardDeviation(index, 0));
 376     }
 377 }
 378
 379 }       // namespace
 380
 381 const char DistanceInfo::name[]             = "distance";
 382 const char DistanceInfo::shortDescription[] =
 383     "Calculate distances between pairs of positions";
 384
 385 TrajectoryAnalysisModulePointer DistanceInfo::create()
 386 {
 387     return TrajectoryAnalysisModulePointer(new Distance);
 388 }
 389
 390 } // namespace analysismodules
 391
 392 } // namespace gmx