MSSpectrum.h

Go to the documentation of this file.

// Copyright (c) 2002-present, OpenMS Inc. -- EKU Tuebingen, ETH Zurich, and FU Berlin
// SPDX-License-Identifier: BSD-3-Clause
//
// --------------------------------------------------------------------------
// $Maintainer: Timo Sachsenberg$
// $Authors: Marc Sturm $
// --------------------------------------------------------------------------
 
#pragma once
 
#include <OpenMS/KERNEL/Peak1D.h>
#include <OpenMS/KERNEL/PeakContainerSort.h>
#include <OpenMS/METADATA/SpectrumSettings.h>
#include <OpenMS/KERNEL/RangeManager.h>
#include <OpenMS/METADATA/DataArrays.h>
#include <OpenMS/METADATA/MetaInfoDescription.h>
 
#include <numeric>
 
namespace OpenMS
{
  enum class DriftTimeUnit;
  class OPENMS_DLLAPI MSSpectrum final :
    private std::vector<Peak1D>,
    public RangeManagerContainer<RangeMZ, RangeIntensity, RangeMobility>,
    public SpectrumSettings
  {
public:
 
    struct OPENMS_DLLAPI RTLess
    {
      bool operator()(const MSSpectrum& a, const MSSpectrum& b) const;
    };
    struct OPENMS_DLLAPI IMLess {
      bool operator()(const MSSpectrum& a, const MSSpectrum& b) const;
    };
 
    struct Chunk {
      Size start; 
      Size end; 
      bool is_sorted; 
      Chunk(Size p_start, Size p_end, bool p_sorted) : start(p_start), end(p_end), is_sorted(p_sorted)
      {
      }
    };
 
    struct Chunks {
      public:
        Chunks(const MSSpectrum& s) : spec_(s) {}
        void add(bool is_sorted)
        {
          chunks_.emplace_back((chunks_.empty() ? 0 : chunks_.back().end), spec_.size(), is_sorted);
        }
        std::vector<Chunk>& getChunks()
        {
          return chunks_;
        }
      private:
        std::vector<Chunk> chunks_;
        const MSSpectrum& spec_;
    };
 
 
 
    typedef OpenMS::Peak1D PeakType;
    typedef typename PeakType::CoordinateType CoordinateType;
    typedef std::vector<PeakType> ContainerType;
    typedef RangeManagerContainer<RangeMZ, RangeIntensity, RangeMobility> RangeManagerContainerType;
    typedef RangeManager<RangeMZ, RangeIntensity, RangeMobility> RangeManagerType;
    typedef OpenMS::DataArrays::FloatDataArray FloatDataArray ;
    typedef std::vector<FloatDataArray> FloatDataArrays;
    typedef OpenMS::DataArrays::StringDataArray StringDataArray ;
    typedef std::vector<StringDataArray> StringDataArrays;
    typedef OpenMS::DataArrays::IntegerDataArray IntegerDataArray ;
    typedef std::vector<IntegerDataArray> IntegerDataArrays;
 
 
 
    typedef typename ContainerType::iterator Iterator;
    typedef typename ContainerType::const_iterator ConstIterator;
    typedef typename ContainerType::reverse_iterator ReverseIterator;
    typedef typename ContainerType::const_reverse_iterator ConstReverseIterator;
 
 
    using ContainerType::operator[];
    using ContainerType::begin;
    using ContainerType::rbegin;
    using ContainerType::end;
    using ContainerType::rend;
    using ContainerType::cbegin;
    using ContainerType::cend;
    using ContainerType::resize;
    using ContainerType::size;
    using ContainerType::push_back;
    using ContainerType::emplace_back;
    using ContainerType::pop_back;
    using ContainerType::empty;
    using ContainerType::front;
    using ContainerType::back;
    using ContainerType::reserve;
    using ContainerType::insert;
    using ContainerType::erase;
    using ContainerType::swap;
    using ContainerType::data;
    using ContainerType::shrink_to_fit;
 
    using typename ContainerType::iterator;
    using typename ContainerType::const_iterator;
    using typename ContainerType::size_type;
    using typename ContainerType::value_type;
    using typename ContainerType::reference;
    using typename ContainerType::const_reference;
    using typename ContainerType::pointer;
    using typename ContainerType::difference_type;
 
 
 
    MSSpectrum();
 
    MSSpectrum(const std::initializer_list<Peak1D>& init);
 
    MSSpectrum(const MSSpectrum& source);
 
    MSSpectrum(MSSpectrum&&) = default;
 
    ~MSSpectrum() = default;
 
    MSSpectrum& operator=(const MSSpectrum& source);
 
    MSSpectrum& operator=(MSSpectrum&&) & = default;
 
    MSSpectrum& operator=(const SpectrumSettings & source);
 
    bool operator==(const MSSpectrum& rhs) const;
 
    bool operator!=(const MSSpectrum& rhs) const
    {
      return !(operator==(rhs));
    }
 
    // Docu in base class (RangeManager)
    void updateRanges() override;
 
    double getRT() const;
 
    void setRT(double rt);
 
    double getDriftTime() const;
 
    void setDriftTime(double dt);
 
    DriftTimeUnit getDriftTimeUnit() const;
 
    std::string getDriftTimeUnitAsString() const;
 
    void setDriftTimeUnit(DriftTimeUnit dt);
 
    UInt getMSLevel() const;
 
    void setMSLevel(UInt ms_level);
 
    const std::string& getName() const;
 
    void setName(const std::string& name);
 
 
    const FloatDataArrays& getFloatDataArrays() const;
 
    FloatDataArrays& getFloatDataArrays()
    {
      return float_data_arrays_;
    }
 
    void setFloatDataArrays(const FloatDataArrays& fda);
 
    const StringDataArrays& getStringDataArrays() const;
 
    StringDataArrays& getStringDataArrays();
 
    void setStringDataArrays(const StringDataArrays& sda);
 
    const IntegerDataArrays& getIntegerDataArrays() const;
 
    IntegerDataArrays& getIntegerDataArrays();
 
    void setIntegerDataArrays(const IntegerDataArrays& ida);
 
 
 
    void sortByIntensity(bool reverse = false);
 
    void sortByPosition();
 
    void sortByIonMobility();
 
    void sortByPositionPresorted(const std::vector<Chunk>& chunks);
 
    bool isSorted() const;
 
    bool isSortedByIM() const;
 
    template<class Predicate>
    bool isSorted(const Predicate& lambda) const
    {
      auto value_2_index_wrapper = [this, &lambda](const value_type& value1, const value_type& value2) {
        // translate values into indices (this relies on no copies being made!)
        const Size index1 = (&value1) - (&this->front());
        const Size index2 = (&value2) - (&this->front());
        // just make sure the pointers above are actually pointing to a Peak inside our container
        assert(index1 < this->size()); 
        assert(index2 < this->size());
        return lambda(index1, index2);
      }; 
      return std::is_sorted(this->begin(), this->end(), value_2_index_wrapper);
    }
 
    template<class Predicate>
    void sort(const Predicate& lambda)
    {
      // Validate up front, before running the (possibly data-array-indexing) predicate, so a
      // mis-sized array throws instead of being read out of bounds during the sort.
      checkDataArraySizes_();
      // The sorted indices are a permutation of [0, size): in range and duplicate-free by
      // construction, so the unchecked path is safe and skips the redundant bounds re-scan.
      selectUnchecked(Internal::stableSortedIndices(this->size(), lambda));
    }
 
 
 
    Size findNearest(CoordinateType mz) const;
 
    Int findNearest(CoordinateType mz, CoordinateType tolerance) const;
 
    Int findNearest(CoordinateType mz, CoordinateType tolerance_left, CoordinateType tolerance_right) const;
 
    Int findHighestInWindow(CoordinateType mz, CoordinateType tolerance_left, CoordinateType tolerance_right) const;
 
    Iterator MZBegin(CoordinateType mz);
 
    Iterator MZBegin(Iterator begin, CoordinateType mz, Iterator end);
 
    Iterator MZEnd(CoordinateType mz);
 
    Iterator MZEnd(Iterator begin, CoordinateType mz, Iterator end);
 
    ConstIterator MZBegin(CoordinateType mz) const;
 
    ConstIterator MZBegin(ConstIterator begin, CoordinateType mz, ConstIterator end) const;
 
    ConstIterator MZEnd(CoordinateType mz) const;
 
    ConstIterator MZEnd(ConstIterator begin, CoordinateType mz, ConstIterator end) const;
 
    Iterator PosBegin(CoordinateType mz);
 
    Iterator PosBegin(Iterator begin, CoordinateType mz, Iterator end);
 
    ConstIterator PosBegin(CoordinateType mz) const;
 
    ConstIterator PosBegin(ConstIterator begin, CoordinateType mz, ConstIterator end) const;
 
    Iterator PosEnd(CoordinateType mz);
 
    Iterator PosEnd(Iterator begin, CoordinateType mz, Iterator end);
 
    ConstIterator PosEnd(CoordinateType mz) const;
 
    ConstIterator PosEnd(ConstIterator begin, CoordinateType mz, ConstIterator end) const;
 
    bool containsIMData() const;
 
    std::pair<Size, DriftTimeUnit> getIMData() const;
    
 
    std::pair<DriftTimeUnit, std::vector<float>> maybeGetIMData() const;
    
 
 
    void clear(bool clear_meta_data);
 
    MSSpectrum& select(const std::vector<Size>& indices);
 
    MSSpectrum& selectUnchecked(const std::vector<Size>& indices);
 
 
    SpectrumSettings::SpectrumType getType(const bool query_data) const;
    using SpectrumSettings::getType; // expose base class function
 
    ConstIterator getBasePeak() const;
 
    Iterator getBasePeak();
 
    PeakType::IntensityType calculateTIC() const;
 
    enum class RasterAggregation
    {
      SUM,  
      MAX   
    };
 
    void rasterizeIMFrame(
      float* output,
      Size im_bins,
      Size mz_bins,
      CoordinateType min_im,
      CoordinateType max_im,
      CoordinateType min_mz,
      CoordinateType max_mz,
      RasterAggregation aggregation = RasterAggregation::SUM) const;
 
protected:
    void checkDataArraySizes_() const;
 
    double retention_time_ = -1;
 
    double drift_time_ = -1;
 
    DriftTimeUnit drift_time_unit_ = DriftTimeUnit::NONE;
 
    UInt ms_level_ = 1;
 
    std::string name_;
 
    FloatDataArrays float_data_arrays_;
 
    StringDataArrays string_data_arrays_;
 
    IntegerDataArrays integer_data_arrays_;
  };
 
  inline std::ostream& operator<<(std::ostream& os, const MSSpectrum& spec)
  {
    os << "-- MSSPECTRUM BEGIN --" << std::endl;
 
    // spectrum settings
    os << static_cast<const SpectrumSettings&>(spec);
 
    // peaklist
    for (MSSpectrum::ConstIterator it = spec.begin(); it != spec.end(); ++it)
    {
      os << *it << std::endl;
    }
 
    os << "-- MSSPECTRUM END --" << std::endl;
    return os;
  }
 
} // namespace OpenMS