OpenMS  2.4.0
DTA2DFile.h
Go to the documentation of this file.
1 // --------------------------------------------------------------------------
2 // OpenMS -- Open-Source Mass Spectrometry
3 // --------------------------------------------------------------------------
4 // Copyright The OpenMS Team -- Eberhard Karls University Tuebingen,
5 // ETH Zurich, and Freie Universitaet Berlin 2002-2018.
6 //
7 // This software is released under a three-clause BSD license:
8 // * Redistributions of source code must retain the above copyright
9 // notice, this list of conditions and the following disclaimer.
10 // * Redistributions in binary form must reproduce the above copyright
11 // notice, this list of conditions and the following disclaimer in the
12 // documentation and/or other materials provided with the distribution.
13 // * Neither the name of any author or any participating institution
14 // may be used to endorse or promote products derived from this software
15 // without specific prior written permission.
16 // For a full list of authors, refer to the file AUTHORS.
17 // --------------------------------------------------------------------------
18 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
19 // AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20 // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21 // ARE DISCLAIMED. IN NO EVENT SHALL ANY OF THE AUTHORS OR THE CONTRIBUTING
22 // INSTITUTIONS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
23 // EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
24 // PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
25 // OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
26 // WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
27 // OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
28 // ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29 //
30 // --------------------------------------------------------------------------
31 // $Maintainer: Timo Sachsenberg $
32 // $Authors: Marc Sturm $
33 // --------------------------------------------------------------------------
34 
35 #pragma once
36 
41 
42 #include <fstream>
43 #include <iostream>
44 
45 namespace OpenMS
46 {
63  class OPENMS_DLLAPI DTA2DFile :
64  public ProgressLogger
65  {
66 private:
68 
69 public:
70 
73  DTA2DFile();
76  ~DTA2DFile();
78 
80  PeakFileOptions& getOptions();
81 
83  const PeakFileOptions& getOptions() const;
84 
94  template <typename MapType>
95  void load(const String& filename, MapType& map)
96  {
97  startProgress(0, 0, "loading DTA2D file");
98 
99  //try to open file
100  std::ifstream is(filename.c_str());
101  if (!is)
102  {
103  throw Exception::FileNotFound(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, filename);
104  }
105 
106  map.reset();
107 
108  //set DocumentIdentifier
109  map.setLoadedFileType(filename);
110  map.setLoadedFilePath(filename);
111 
112  // temporary variables to store the data in
113  std::vector<String> strings(3);
114  typename MapType::SpectrumType spec;
115  spec.setRT(-1.0); //to make sure the first RT is different from the the initialized value
117  double rt(0.0);
118  char delimiter;
119 
120  // default dimension of the data
121  Size rt_dim = 0;
122  Size mz_dim = 1;
123  Size int_dim = 2;
124 
125  //RT unit (default is seconds)
126  bool time_in_minutes = false;
127 
128  // string to store the current line in
129  String line;
130 
131  // native ID (numbers from 0)
132  UInt native_id = 0;
133 
134  // line number counter
135  Size line_number = 0;
136 
137  while (getline(is, line, '\n'))
138  {
139  ++line_number;
140  line.trim();
141 
142  if (line.empty()) continue;
143 
144  //test which delimiter is used in the line
145  if (line.has('\t'))
146  {
147  delimiter = '\t';
148  }
149  else
150  {
151  delimiter = ' ';
152  }
153 
154  //is header line
155  if (line.hasPrefix("#"))
156  {
157  line = line.substr(1).trim();
158  line.split(delimiter, strings);
159 
160  // flags to check if dimension is set correctly
161  bool rt_set = false;
162  bool mz_set = false;
163  bool int_set = false;
164 
165  //assign new order
166  for (Size i = 0; i < 3; ++i)
167  {
168  if (strings[i] == "RT" || strings[i] == "RETENTION_TIME" || strings[i] == "MASS-TO-CHARGE" || strings[i] == "IT" || strings[i] == "INTENSITY")
169  {
170  std::cerr << "Warning: This file contains the deprecated keyword '" << strings[i] << "'." << "\n";
171  std::cerr << " Please use only the new keywords SEC/MIN, MZ, INT." << "\n";
172  }
173  if ((strings[i] == "SEC" || strings[i] == "RT" || strings[i] == "RETENTION_TIME") && rt_set == false)
174  {
175  rt_dim = i;
176  rt_set = true;
177  }
178  else if ((strings[i] == "MIN") && rt_set == false)
179  {
180  rt_dim = i;
181  rt_set = true;
182  time_in_minutes = true;
183  }
184  else if ((strings[i] == "MZ" || strings[i] == "MASS-TO-CHARGE") && mz_set == false)
185  {
186  mz_dim = i;
187  mz_set = true;
188  }
189  else if ((strings[i] == "INT" || strings[i] == "IT" || strings[i] == "INTENSITY") && int_set == false)
190  {
191  int_dim = i;
192  int_set = true;
193  }
194  else
195  {
196  throw Exception::ParseError(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, "Misformatted header line!", filename);
197  }
198  }
199  continue;
200  }
201 
202  try
203  {
204  line.split(delimiter, strings);
205  if (strings.size() != 3)
206  {
207  throw Exception::ParseError(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, std::string("Bad data line (" + String(line_number) + "): \"") + line + "\" (got " + String(strings.size()) + ", expected 3 entries)", filename);
208  }
209  p.setIntensity(strings[int_dim].toFloat());
210  p.setMZ(strings[mz_dim].toDouble());
211  rt = (strings[rt_dim].toDouble()) * (time_in_minutes ? 60.0 : 1.0);
212  }
213  // conversion to double or something else could have gone wrong
214  catch (Exception::BaseException& /*e*/)
215  {
216  throw Exception::ParseError(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, std::string("Bad data line (" + String(line_number) + "): \"") + line + "\"", filename);
217  }
218 
219  // Retention time changed -> new Spectrum
220  if (fabs(rt - spec.getRT()) > 0.0001)
221  {
222  if (spec.size() != 0
223  &&
224  (!options_.hasRTRange() || options_.getRTRange().encloses(DPosition<1>(spec.getRT())))) // RT restriction fulfilled
225  {
226  map.addSpectrum(spec);
227  }
228  setProgress(0);
229  spec.clear(true);
230  spec.setRT(rt);
231  spec.setNativeID(String("index=") + native_id);
232  ++native_id;
233  }
234 
235  //Skip peaks with invalid m/z or intensity value
236  if (
237  (!options_.hasMZRange() || options_.getMZRange().encloses(DPosition<1>(p.getMZ())))
238  &&
239  (!options_.hasIntensityRange() || options_.getIntensityRange().encloses(DPosition<1>(p.getIntensity())))
240  )
241  {
242  spec.push_back(p);
243  }
244  }
245 
246  // add last Spectrum
247  if (
248  spec.size() != 0
249  &&
250  (!options_.hasRTRange() || options_.getRTRange().encloses(DPosition<1>(spec.getRT()))) // RT restriction fulfilled
251  )
252  {
253  map.addSpectrum(spec);
254  }
255 
256  is.close();
257  endProgress();
258  }
259 
268  template <typename MapType>
269  void store(const String& filename, const MapType& map) const
270  {
271  startProgress(0, map.size(), "storing DTA2D file");
272 
273  std::ofstream os(filename.c_str());
274  if (!os)
275  {
276  throw Exception::UnableToCreateFile(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, filename);
277  }
278 
279  // write header
280  os << "#SEC\tMZ\tINT\n";
281 
282  // Iterate over all peaks of each spectrum and
283  // write one line for each peak of the spectrum.
284  UInt count = 0;
285  for (typename MapType::const_iterator spec = map.begin(); spec != map.end(); ++spec)
286  {
287  setProgress(count++);
288  for (typename MapType::SpectrumType::ConstIterator it = spec->begin(); it != spec->end(); ++it)
289  {
290  // Write rt, m/z and intensity.
291  os << precisionWrapper(spec->getRT()) << "\t" << precisionWrapper(it->getPos()) << "\t" << precisionWrapper(it->getIntensity()) << "\n";
292  }
293 
294  }
295  os.close();
296  endProgress();
297  }
298 
307  template <typename MapType>
308  void storeTIC(const String& filename, const MapType& map) const
309  {
310  startProgress(0, map.size(), "storing DTA2D file");
311 
312  std::ofstream os(filename.c_str());
313  if (!os)
314  {
315  throw Exception::UnableToCreateFile(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, filename);
316  }
317 
318  // write header (Always MZ=0 for chromatograms in DTA2D.)
319  os << "#SEC\tMZ\tINT\n";
320 
321  typename MapType::ChromatogramType TIC = map.getTIC();
322  for (typename MapType::ChromatogramType::ConstIterator it = TIC.begin(); it != TIC.end(); ++it)
323  {
324  // write rt and intensity.
325  os << precisionWrapper(it->getRT()) << "\t" << precisionWrapper(0) << "\t" << precisionWrapper(it->getIntensity()) << "\n";
326  }
327 
328  os.close();
329  endProgress();
330  }
331 
332  };
333 
334 } // namespace OpenMS
335 
DTA2D File adapter.
Definition: DTA2DFile.h:63
bool has(Byte byte) const
true if String contains the byte, false otherwise
A more convenient string class.
Definition: String.h:57
bool encloses(const PositionType &position) const
Checks whether this range contains a certain point.
Definition: DRange.h:171
void reset()
Resets all internal values.
void addSpectrum(const MSSpectrum &spectrum)
adds a spectrum to the list
The representation of a chromatogram.
Definition: MSChromatogram.h:54
CoordinateType getMZ() const
Non-mutable access to m/z.
Definition: Peak1D.h:113
unsigned int UInt
Unsigned integer type.
Definition: Types.h:94
const DRange< 1 > & getMZRange() const
returns the MZ range
PeakFileOptions options_
Definition: DTA2DFile.h:67
Iterator begin()
Definition: MSExperiment.h:157
Base::const_iterator const_iterator
Definition: MSExperiment.h:125
bool hasRTRange() const
returns true if an RT range has been set
File not found exception.
Definition: Exception.h:523
Size size() const
Definition: MSExperiment.h:127
Main OpenMS namespace.
Definition: FeatureDeconvolution.h:46
void setMZ(CoordinateType mz)
Mutable access to m/z.
Definition: Peak1D.h:119
void setIntensity(IntensityType intensity)
Mutable access to the data point intensity (height)
Definition: Peak1D.h:110
const DRange< 1 > & getRTRange() const
returns the RT range
void setLoadedFileType(const String &file_name)
set the file_type according to the type of the file loaded from (see FileHandler::Type) preferably do...
bool hasMZRange() const
returns true if an MZ range has been set
const DRange< 1 > & getIntensityRange() const
returns the intensity range
String substr(size_t pos=0, size_t n=npos) const
Wrapper for the STL substr() method. Returns a String object with its contents initialized to a subst...
Iterator end()
Definition: MSExperiment.h:167
The representation of a 1D spectrum.
Definition: MSSpectrum.h:66
void store(const String &filename, const MapType &map) const
Stores a map in a DTA2D file.
Definition: DTA2DFile.h:269
const PrecisionWrapper< FloatingPointType > precisionWrapper(const FloatingPointType rhs)
Wrapper function that sets the appropriate precision for output temporarily. The original precision i...
Definition: PrecisionWrapper.h:96
String & trim()
removes whitespaces (space, tab, line feed, carriage return) at the beginning and the end of the stri...
A 1-dimensional raw data point or peak.
Definition: Peak1D.h:54
Exception base class.
Definition: Exception.h:89
void clear(bool clear_meta_data)
Clears all data and meta data.
void setRT(double rt)
Sets the absolute retention time (in seconds)
In-Memory representation of a mass spectrometry experiment.
Definition: MSExperiment.h:77
const MSChromatogram getTIC() const
returns the total ion chromatogram (TIC)
std::vector< SpectrumType >::const_iterator ConstIterator
Non-mutable iterator.
Definition: MSExperiment.h:113
bool hasPrefix(const String &string) const
true if String begins with string, false otherwise
size_t Size
Size type e.g. used as variable which can hold result of size()
Definition: Types.h:127
bool hasIntensityRange() const
returns true if an intensity range has been set
Base class for all classes that want to report their progress.
Definition: ProgressLogger.h:54
Unable to create file exception.
Definition: Exception.h:636
void setLoadedFilePath(const String &file_name)
set the file_name_ according to absolute path of the file loaded from preferably done whilst loading ...
void setNativeID(const String &native_id)
sets the native identifier for the spectrum, used by the acquisition software.
Options for loading files containing peak data.
Definition: PeakFileOptions.h:47
void storeTIC(const String &filename, const MapType &map) const
Stores the TIC of a map in a DTA2D file.
Definition: DTA2DFile.h:308
double getRT() const
bool split(const char splitter, std::vector< String > &substrings, bool quote_protect=false) const
Splits a string into substrings using splitter as delimiter.
void load(const String &filename, MapType &map)
Loads a map from a DTA2D file.
Definition: DTA2DFile.h:95
Parse Error exception.
Definition: Exception.h:622
IntensityType getIntensity() const
Definition: Peak1D.h:108