src/gromacs/analysisdata/datastorage.h

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014,2019, 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 /*! \libinternal \file
  36  * \brief
  37  * Declares gmx::AnalysisDataStorage.
  38  *
  39  * \author Teemu Murtola <teemu.murtola@gmail.com>
  40  * \inlibraryapi
  41  * \ingroup module_analysisdata
  42  */
  43 #ifndef GMX_ANALYSISDATA_DATASTORAGE_H
  44 #define GMX_ANALYSISDATA_DATASTORAGE_H
  45
  46 #include <vector>
  47
  48 #include "gromacs/analysisdata/dataframe.h"
  49 #include "gromacs/utility/classhelpers.h"
  50 #include "gromacs/utility/gmxassert.h"
  51 #include "gromacs/utility/real.h"
  52
  53 namespace gmx
  54 {
  55
  56 class AbstractAnalysisData;
  57 class AnalysisDataFrameHeader;
  58 class AnalysisDataFrameRef;
  59 class AnalysisDataModuleManager;
  60 class AnalysisDataParallelOptions;
  61
  62 class AnalysisDataStorage;
  63
  64 namespace internal
  65 {
  66 class AnalysisDataStorageImpl;
  67 class AnalysisDataStorageFrameData;
  68 } // namespace internal
  69
  70 /*! \libinternal \brief
  71  * Allows assigning values for a data frame in AnalysisDataStorage.
  72  *
  73  * This class implements the necessary methods to add new data into the
  74  * storage.  AnalysisDataStorage::startFrame() returns an object of this type,
  75  * which can be used to add one or more point sets to that data frame.
  76  * When all data has been added, finishFrame() needs to be called.
  77  *
  78  * \inlibraryapi
  79  * \ingroup module_analysisdata
  80  */
  81 class AnalysisDataStorageFrame
  82 {
  83 public:
  84     /*! \brief Frees the frame object.
  85      *
  86      * Should not be called outside AnalysisDataStorage.
  87      */
  88     ~AnalysisDataStorageFrame();
  89
  90     /*! \brief
  91      * Select data set that all other methods operate on.
  92      *
  93      * \param[in] index  Zero-based data set index to select.
  94      *
  95      * With multipoint data, a single point set can only contain values in
  96      * a single data set.
  97      * With non-multipoint data, arbitrary sequences of selectDataSet() and
  98      * setValue() are supported.  The full frame is notified to the modules
  99      * once it is finished.
 100      *
 101      * Does not throw.
 102      */
 103     void selectDataSet(int index);
 104
 105     //! Returns number of columns for the frame.
 106     int columnCount() const { return columnCount_; }
 107
 108     /*! \brief
 109      * Sets value for a column.
 110      *
 111      * \param[in] column  Zero-based column index.
 112      * \param[in] value   Value to set for the column.
 113      * \param[in] bPresent Present flag to set for the column.
 114      *
 115      * If called multiple times for a column (within one point set for
 116      * multipoint data), old values are overwritten.
 117      *
 118      * Does not throw.
 119      */
 120     void setValue(int column, real value, bool bPresent = true)
 121     {
 122         GMX_ASSERT(column >= 0 && column < columnCount(), "Invalid column index");
 123         values_[currentOffset_ + column].setValue(value, bPresent);
 124         bPointSetInProgress_ = true;
 125     }
 126     /*! \brief
 127      * Sets value for a column.
 128      *
 129      * \param[in] column  Zero-based column index.
 130      * \param[in] value   Value to set for the column.
 131      * \param[in] error   Error estimate to set for the column.
 132      * \param[in] bPresent Present flag to set for the column.
 133      *
 134      * If called multiple times for a column (within one point set for
 135      * multipoint data), old values are overwritten.
 136      *
 137      * Does not throw.
 138      */
 139     void setValue(int column, real value, real error, bool bPresent = true)
 140     {
 141         GMX_ASSERT(column >= 0 && column < columnCount(), "Invalid column index");
 142         values_[currentOffset_ + column].setValue(value, error, bPresent);
 143         bPointSetInProgress_ = true;
 144     }
 145     /*! \brief
 146      * Access value for a column.
 147      *
 148      * \param[in] column  Zero-based column index.
 149      *
 150      * Should only be called after the column value has been set using
 151      * setValue(); assigning a value to \c value(i) does not mark the
 152      * column as set.
 153      *
 154      * Does not throw.
 155      */
 156     real& value(int column)
 157     {
 158         GMX_ASSERT(column >= 0 && column < columnCount(), "Invalid column index");
 159         return values_[currentOffset_ + column].value();
 160     }
 161     /*! \brief
 162      * Access value for a column.
 163      *
 164      * \param[in] column  Zero-based column index.
 165      *
 166      * Should only be called after the column value has been set using
 167      * setValue().
 168      *
 169      * Does not throw.
 170      */
 171     real value(int column) const
 172     {
 173         GMX_ASSERT(column >= 0 && column < columnCount(), "Invalid column index");
 174         return values_[currentOffset_ + column].value();
 175     }
 176     /*! \brief
 177      * Mark point set as finished for multipoint data.
 178      *
 179      * Must be called after each point set for multipoint data, including
 180      * the last (i.e., no values must be set between the last call to this
 181      * method and AnalysisDataStorage::finishFrame()).
 182      * Must not be called for non-multipoint data.
 183      *
 184      * After this method has been called, all values appear as not set.
 185      *
 186      * May call AnalysisDataModuleManager::notifyPointsAdd() and
 187      * AnalysisDataModuleManager::notifyParallelPointsAdd(), and may throw
 188      * any exception these methods throw.
 189      */
 190     void finishPointSet();
 191     /*! \brief
 192      * Finish storing a frame.
 193      *
 194      * Must be called exactly once for each frame returned by startFrame(),
 195      * after the corresponding call.
 196      * The frame object must not be accessed after the call.
 197      *
 198      * Calls notification methods in AnalysisDataModuleManager, and may
 199      * throw any exceptions these methods throw.
 200      */
 201     void finishFrame();
 202
 203 private:
 204     /*! \brief
 205      * Create a new storage frame.
 206      *
 207      * \param[in] data  Data object for which the frame is for
 208      *      (used for data set and column counts).
 209      */
 210     explicit AnalysisDataStorageFrame(const AbstractAnalysisData& data);
 211
 212     //! Clear all column values from the frame.
 213     void clearValues();
 214
 215     //! Implementation data.
 216     internal::AnalysisDataStorageFrameData* data_;
 217     //! Values for the currently in-progress point set.
 218     std::vector<AnalysisDataValue> values_;
 219
 220     //! Index of the currently active dataset.
 221     int currentDataSet_;
 222     //! Offset of the first value in \a values_ for the current data set.
 223     int currentOffset_;
 224     //! Number of columns in the current data set.
 225     int columnCount_;
 226
 227     //! Whether any values have been set in the current point set.
 228     bool bPointSetInProgress_;
 229
 230     //! Needed for access to the constructor.
 231     friend class internal::AnalysisDataStorageImpl;
 232     //! Needed for managing the frame the object points to.
 233     friend class internal::AnalysisDataStorageFrameData;
 234
 235     GMX_DISALLOW_COPY_AND_ASSIGN(AnalysisDataStorageFrame);
 236 };
 237
 238 /*! \libinternal \brief
 239  * Helper class that implements storage of data.
 240  *
 241  * This class implements a standard way of storing data to avoid implementing
 242  * storage in each class derived from AbstractAnalysisData separately.
 243  * To use this class in a class derived from AbstractAnalysisData, a member
 244  * variable of this type should be declared and the pure virtual methods
 245  * forwarded to frameCount(), tryGetDataFrame() and requestStorage().
 246  * Storage properties should be set up, and then startDataStorage() or
 247  * startParallelDataStorage() called.
 248  * New frames can then be added using startFrame(), currentFrame(),
 249  * finishFrame(), and finishFrameSerial() methods (the last is only necessary
 250  * if startParallelDataStorage() is used).  When all frames are ready,
 251  * finishDataStorage() must be called.  These methods (and
 252  * AnalysisDataStorageFrame::finishPointSet()) take the responsibility of
 253  * calling all the notification methods in AnalysisDataModuleManager,
 254  *
 255  * \todo
 256  * Proper multi-threaded implementation.
 257  *
 258  * \inlibraryapi
 259  * \ingroup module_analysisdata
 260  */
 261 class AnalysisDataStorage
 262 {
 263 public:
 264     //! Constructs a storage object.
 265     AnalysisDataStorage();
 266     ~AnalysisDataStorage();
 267
 268     /*! \brief
 269      * Returns the number of ready frames.
 270      *
 271      * This method is designed such that calls to
 272      * AbstractAnalysisData::frameCount() can be directly forwarded to this
 273      * method.  See that method for more documentation.
 274      *
 275      * If this method returns N, this means that the first N frames have
 276      * all been finished.
 277      *
 278      * \see AbstractAnalysisData::frameCount()
 279      */
 280     int frameCount() const;
 281     /*! \brief
 282      * Implements access to data frames.
 283      *
 284      * This method is designed such that calls to
 285      * AbstractAnalysisData::tryGetDataFrameInternal() can be directly
 286      * forwarded to this method.  See that method for more documentation.
 287      *
 288      * A valid reference for a frame will be returned after finishFrame()
 289      * has been called for that frame.
 290      *
 291      * \see AbstractAnalysisData::tryGetDataFrameInternal()
 292      */
 293     AnalysisDataFrameRef tryGetDataFrame(int index) const;
 294     /*! \brief
 295      * Implements storage requests.
 296      *
 297      * This method is designed such that calls to
 298      * AbstractAnalysisData::requestStorageInternal() can be directly
 299      * forwarded to this method.  See that method for more documentation.
 300      *
 301      * \see AbstractAnalysisData::requestStorageInternal()
 302      */
 303     bool requestStorage(int nframes);
 304
 305     /*! \brief
 306      * Start storing data.
 307      *
 308      * \param[in] data    AbstractAnalysisData object containing this
 309      *      storage.
 310      * \param     modules Module manager for \p data.
 311      * \exception std::bad_alloc if storage allocation fails.
 312      *
 313      * Typically called as \c startDataStorage(this, &moduleManager())
 314      * from a member of \p data when the data is ready to be started.
 315      * The storage object will take responsibility of calling all
 316      * module notification methods in AnalysisDataModuleManager using
 317      * \p modules.
 318      *
 319      * Lifetime of \p data and \p modules must exceed the lifetime of the
 320      * storage object
 321      * (typically, the storage object will be a member in \p data).
 322      *
 323      * Calls AnalysisDataModuleManager::notifyDataStart(), and throws any
 324      * exceptions this method throws.
 325      */
 326     void startDataStorage(AbstractAnalysisData* data, AnalysisDataModuleManager* modules);
 327     /*! \brief
 328      * Start storing data in parallel.
 329      *
 330      * \param[in] data    AbstractAnalysisData object containing this
 331      *      storage.
 332      * \param[in] options Parallelization options to use.
 333      * \param     modules Module manager for \p data.
 334      * \exception std::bad_alloc if storage allocation fails.
 335      *
 336      * Should be called instead of startDataStorage() if the data will be
 337      * produced in parallel.  Works as startDataStorage(), but additionally
 338      * initializes the storage and the attached modules to prepare for
 339      * out-of-order data frames.
 340      *
 341      * Calls AnalysisDataModuleManager::notifyParallelDataStart(), and
 342      * throws any exceptions this method throws.
 343      */
 344     void startParallelDataStorage(AbstractAnalysisData*              data,
 345                                   AnalysisDataModuleManager*         modules,
 346                                   const AnalysisDataParallelOptions& options);
 347     /*! \brief
 348      * Starts storing a new frame.
 349      *
 350      * \param[in] header  Header for the new frame.
 351      * \retval  Frame object corresponding to the started frame.
 352      * \exception std::bad_alloc if storage reallocation fails
 353      *      (only possible if storage of all frames has been requested).
 354      * \exception APIError if frame is too far in the future.
 355      *
 356      * The returned object will be valid until the corresponding
 357      * finishFrame() call.
 358      *
 359      * Must be called exactly once for each frame index.
 360      *
 361      * Currently, the implementation only works if the new frame is not too
 362      * far in the future:
 363      * If \c i is the index of the last frame such that all frames from
 364      * 0, ..., \c i have been finished, then \p header().index() should be
 365      * at most \c parallelizationFactor larger than \c i, where
 366      * parallelizationFactor is the parallelization factor passed to
 367      * setParallelOptions().
 368      * Throws APIError if this constraint is violated.
 369      *
 370      * Calls AnalysisDataModuleManager::notifyFrameStart() (in certain
 371      * cases) and AnalysisDataModuleManager::notifyParallelFrameStart(),
 372      * and throws any exceptions these methods throw.
 373      */
 374     AnalysisDataStorageFrame& startFrame(const AnalysisDataFrameHeader& header);
 375     /*! \brief
 376      * Convenience method to start storing a new frame.
 377      *
 378      * Identical to \c startFrame(AnalysisDataFrameHeader(index, x, dx));
 379      */
 380     AnalysisDataStorageFrame& startFrame(int index, real x, real dx);
 381     /*! \brief
 382      * Obtains a frame object for an in-progress frame.
 383      *
 384      * \param[in] index  Frame index.
 385      * \retval  Frame object corresponding to \p index.
 386      *
 387      * startFrame() should have been called for the frame with index
 388      * \p index, and finishFrame() should not yet have been called.
 389      * Returns the same object as returned by the original startFrame()
 390      * call for the same index.
 391      *
 392      * Does not throw.
 393      */
 394     AnalysisDataStorageFrame& currentFrame(int index);
 395     /*! \brief
 396      * Convenience method for finishing a data frame.
 397      *
 398      * \param[in] index  Frame index.
 399      *
 400      * Identical to \c currentFrame(index).finishFrame().
 401      *
 402      * \see AnalysisDataStorageFrame::finishFrame()
 403      */
 404     void finishFrame(int index);
 405     /*! \brief
 406      * Performs in-order sequential processing for a data frame.
 407      *
 408      * \param[in] index  Frame index.
 409      *
 410      * If startParallelDataStorage() has been called with options that
 411      * indicate parallelism, this method must be called after
 412      * `finishFrame(index)` (or the equivalent call in
 413      * AnalysisDataStorageFrame), such that it is called in the correct
 414      * order sequentially for each frame.
 415      *
 416      * If there is no parallelism, this method does nothing; the equivalent
 417      * processing is done already during finishFrame().
 418      */
 419     void finishFrameSerial(int index);
 420     /*! \brief
 421      * Finishes storing data.
 422      *
 423      * Calls AnalysisDataModuleManager::notifyDataFinish(), and throws any
 424      * exceptions this method throws.
 425      */
 426     void finishDataStorage();
 427
 428 private:
 429     typedef internal::AnalysisDataStorageImpl Impl;
 430
 431     PrivateImplPointer<Impl> impl_;
 432 };
 433
 434 } // namespace gmx
 435
 436 #endif