Provides access to linearly interpolated values (and derivatives) from discrete data points. Values beyond the given range of data points are implicitly taken as zero. More...

#include <OpenMS/MATH/MISC/LinearInterpolation.h>

Public Types
Typedefs
typedef Value	value_type

typedef Key	key_type

typedef std::vector< value_type >	container_type

typedef value_type	ValueType

typedef key_type	KeyType

typedef container_type	ContainerType

Public Member Functions
	LinearInterpolation (KeyType scale=1., KeyType offset=0.)
	Constructors and destructor. More...

	LinearInterpolation (LinearInterpolation const &arg)
	Copy constructor. More...

LinearInterpolation &	operator= (LinearInterpolation const &arg)
	Assignment operator. More...

	~LinearInterpolation ()
	Destructor. More...

Interpolated data
ValueType	value (KeyType arg_pos) const
	Returns the interpolated value. More...

void	addValue (KeyType arg_pos, ValueType arg_value)
	Performs linear resampling. The arg_value is split up and added to the data points around arg_pos. More...

ValueType	derivative (KeyType arg_pos) const
	Returns the interpolated derivative. More...

Discrete (non-interpolated) data
ContainerType &	getData ()
	Returns the internal random access container from which interpolated values are being sampled. More...

ContainerType const &	getData () const
	Returns the internal random access container from which interpolated values are being sampled. More...

template<typename SourceContainer >
void	setData (SourceContainer const &data)
	Assigns data to the internal random access container from which interpolated values are being sampled. More...

bool	empty () const
	Returns `true` if getData() is empty. More...

Transformation
KeyType	key2index (KeyType pos) const
	The transformation from "outside" to "inside" coordinates. More...

KeyType	index2key (KeyType pos) const
	The transformation from "inside" to "outside" coordinates. More...

KeyType const &	getScale () const
	Accessor. "Scale" is the difference (in "outside" units) between consecutive entries in "Data". More...

void	setScale (KeyType const &scale)
	Accessor. "Scale" is the difference (in "outside" units) between consecutive entries in "Data". More...

KeyType const &	getOffset () const
	Accessor. "Offset" is the point (in "outside" units) which corresponds to "Data[0]". More...

void	setOffset (KeyType const &offset)
	Accessor. "Offset" is the point (in "outside" units) which corresponds to "Data[0]". More...

void	setMapping (KeyType const &scale, KeyType const &inside, KeyType const &outside)
	Specifies the mapping from "outside" to "inside" coordinates by the following data: More...

void	setMapping (KeyType const &inside_low, KeyType const &outside_low, KeyType const &inside_high, KeyType const &outside_high)
	Specifies the mapping from "outside" to "inside" coordinates by the following data: More...

KeyType const &	getInsideReferencePoint () const
	Accessor. See setMapping(). More...

KeyType const &	getOutsideReferencePoint () const
	Accessor. See setMapping(). More...

KeyType	supportMin () const
	Lower boundary of the support, in "outside" coordinates. More...

KeyType	supportMax () const
	Upper boundary of the support, in "outside" coordinates. More...

Protected Attributes
KeyType	scale_

KeyType	offset_

KeyType	inside_

KeyType	outside_

ContainerType	data_

Detailed Description

template<typename Key = double, typename Value = Key>
class OpenMS::Math::LinearInterpolation< Key, Value >

Provides access to linearly interpolated values (and derivatives) from discrete data points. Values beyond the given range of data points are implicitly taken as zero.

The input is just a vector of values ("Data"). These are interpreted as the y-coordinates at the x-coordinate positions 0,...,data_.size-1.

The interpolated data can also be scaled and shifted in the x-dimension by an affine mapping. That is, we have "inside" and "outside" x-coordinates. The affine mapping can be specified in two ways:

using setScale() and setOffset(),
using setMapping()

By default the identity mapping (scale=1, offset=0) is used.

Using the value() and derivative() methods you can sample linearly interpolated values for a given x-coordinate position of the data and the derivative of the data.

See also: BilinearInterpolation

Member Typedef Documentation

◆ container_type

typedef std::vector<value_type> container_type

◆ ContainerType

typedef container_type ContainerType

◆ key_type

typedef Key key_type

◆ KeyType

typedef key_type KeyType

◆ value_type

typedef Value value_type

◆ ValueType

typedef value_type ValueType

Constructor & Destructor Documentation

◆ LinearInterpolation() [1/2]

LinearInterpolation	(	KeyType	scale = `1.`,
		KeyType	offset = `0.`
	)

inline

Constructors and destructor.

The first argument is the scale which is applied to the arguments of value() and derivative() before looking up the interpolated values in the container. The second argument is the offset, which is subtracted before everything else.

◆ LinearInterpolation() [2/2]

LinearInterpolation ( LinearInterpolation< Key, Value > const & arg )

inline

Copy constructor.

◆ ~LinearInterpolation()

~LinearInterpolation ( )

inline

Destructor.

Member Function Documentation

◆ addValue()

void addValue	(	KeyType	arg_pos,
		ValueType	arg_value
	)

inline

Performs linear resampling. The arg_value is split up and added to the data points around arg_pos.

◆ derivative()

ValueType derivative ( KeyType arg_pos ) const

inline

Returns the interpolated derivative.

Please drop me (= the maintainer) a message if you are using this.

◆ empty()

bool empty ( ) const

inline

Returns true if getData() is empty.

Referenced by LinearInterpolation< double >::supportMin().

◆ getData() [1/2]

ContainerType& getData ( )

inline

Returns the internal random access container from which interpolated values are being sampled.

◆ getData() [2/2]

ContainerType const& getData ( ) const

inline

Returns the internal random access container from which interpolated values are being sampled.

◆ getInsideReferencePoint()

KeyType const& getInsideReferencePoint ( ) const

inline

Accessor. See setMapping().

◆ getOffset()

KeyType const& getOffset ( ) const

inline

Accessor. "Offset" is the point (in "outside" units) which corresponds to "Data[0]".

Referenced by MaxLikeliFitter1D::fitOffset_().

◆ getOutsideReferencePoint()

KeyType const& getOutsideReferencePoint ( ) const

inline

Accessor. See setMapping().

◆ getScale()

KeyType const& getScale ( ) const

inline

Accessor. "Scale" is the difference (in "outside" units) between consecutive entries in "Data".

◆ index2key()

KeyType index2key ( KeyType pos ) const

inline

The transformation from "inside" to "outside" coordinates.

Referenced by LinearInterpolation< double >::supportMax(), and LinearInterpolation< double >::supportMin().

◆ key2index()

KeyType key2index ( KeyType pos ) const

inline

The transformation from "outside" to "inside" coordinates.

Referenced by LinearInterpolation< double >::addValue(), LinearInterpolation< double >::derivative(), and LinearInterpolation< double >::value().

◆ operator=()

LinearInterpolation& operator= ( LinearInterpolation< Key, Value > const & arg )

inline

Assignment operator.

◆ setData()

void setData ( SourceContainer const & data )

inline

Assigns data to the internal random access container from which interpolated values are being sampled.

SourceContainer must be assignable to ContainerType.

◆ setMapping() [1/2]

void setMapping	(	KeyType const &	scale,
		KeyType const &	inside,
		KeyType const &	outside
	)

inline

Specifies the mapping from "outside" to "inside" coordinates by the following data:

scale: the difference in outside coordinates between consecutive values in the data vector.
inside and outside: these x-axis positions are mapped onto each other.

For example, when you have a complicated probability distribution which is in fact centered around zero (but you cannot have negative indices in the data vector), then you can arrange things such that inside is the mean of the pre-computed, shifted density values of that distribution and outside is the centroid position of, say, a peak in the real world which you want to model by a scaled and shifted version of the probability distribution.

Referenced by LinearInterpolation< double >::setMapping().

◆ setMapping() [2/2]

void setMapping	(	KeyType const &	inside_low,
		KeyType const &	outside_low,
		KeyType const &	inside_high,
		KeyType const &	outside_high
	)

inline

Specifies the mapping from "outside" to "inside" coordinates by the following data:

inside_low and outside_low: these axis positions are mapped onto each other.
inside_high and outside_high: these axis positions are mapped onto each other.

This four argument version is just a convenience overload for the three argument version, which see.

◆ setOffset()

void setOffset ( KeyType const & offset )

inline

Accessor. "Offset" is the point (in "outside" units) which corresponds to "Data[0]".

Note: Using this invalidates the inside and outside reference points.

◆ setScale()

void setScale ( KeyType const & scale )

inline

Accessor. "Scale" is the difference (in "outside" units) between consecutive entries in "Data".

Note: Using this invalidates the inside and outside reference points.

◆ supportMax()

KeyType supportMax ( ) const

inline

Upper boundary of the support, in "outside" coordinates.

◆ supportMin()

KeyType supportMin ( ) const

inline

Lower boundary of the support, in "outside" coordinates.

Referenced by MaxLikeliFitter1D::fitOffset_().

◆ value()

ValueType value ( KeyType arg_pos ) const

inline

Returns the interpolated value.

Member Data Documentation

◆ data_

ContainerType data_

protected

Referenced by LinearInterpolation< double >::addValue(), LinearInterpolation< double >::derivative(), LinearInterpolation< double >::empty(), LinearInterpolation< double >::getData(), LinearInterpolation< double >::operator=(), LinearInterpolation< double >::setData(), LinearInterpolation< double >::supportMax(), and LinearInterpolation< double >::value().

◆ inside_

KeyType inside_

protected

Referenced by LinearInterpolation< double >::getInsideReferencePoint(), LinearInterpolation< double >::operator=(), and LinearInterpolation< double >::setMapping().

◆ offset_

KeyType offset_

protected

Referenced by LinearInterpolation< double >::getOffset(), LinearInterpolation< double >::index2key(), LinearInterpolation< double >::key2index(), LinearInterpolation< double >::operator=(), LinearInterpolation< double >::setMapping(), and LinearInterpolation< double >::setOffset().

◆ outside_

KeyType outside_

protected

Referenced by LinearInterpolation< double >::getOutsideReferencePoint(), LinearInterpolation< double >::operator=(), and LinearInterpolation< double >::setMapping().

◆ scale_

KeyType scale_

protected

Referenced by LinearInterpolation< double >::getScale(), LinearInterpolation< double >::index2key(), LinearInterpolation< double >::key2index(), LinearInterpolation< double >::operator=(), LinearInterpolation< double >::setMapping(), and LinearInterpolation< double >::setScale().

Public Types

Public Member Functions

Protected Attributes

Detailed Description

template<typename Key = double, typename Value = Key> class OpenMS::Math::LinearInterpolation< Key, Value >

Member Typedef Documentation

◆ container_type

◆ ContainerType

◆ key_type

◆ KeyType

◆ value_type

◆ ValueType

Constructor & Destructor Documentation

◆ LinearInterpolation() [1/2]

◆ LinearInterpolation() [2/2]

◆ ~LinearInterpolation()

Member Function Documentation

◆ addValue()

◆ derivative()

◆ empty()

◆ getData() [1/2]

◆ getData() [2/2]

◆ getInsideReferencePoint()

◆ getOffset()

◆ getOutsideReferencePoint()

◆ getScale()

◆ index2key()

◆ key2index()

◆ operator=()

◆ setData()

◆ setMapping() [1/2]

◆ setMapping() [2/2]

◆ setOffset()

◆ setScale()

◆ supportMax()

◆ supportMin()

◆ value()

Member Data Documentation

◆ data_

◆ inside_

◆ offset_

◆ outside_

◆ scale_

template<typename Key = double, typename Value = Key>
class OpenMS::Math::LinearInterpolation< Key, Value >