Fn_aux

Functions
bool	ecocpak::contains (const ucolvec &A, const u32 n, const u32 i)
ucolvec	ecocpak::complement (const ucolvec &A, const u32 sizeU)
vector< ClassData >	ecocpak::create_class_vector (const mat &samples, const icolvec &labels)
mat	ecocpak::create_samples_matrix (const vector< ClassData > &v)
void	ecocpak::create_samples_matrix (const vector< ClassData > &v, mat &samples, icolvec &labels)
void	ecocpak::create_samples_matrix (const ClassData &first_class, const ClassData &second_class, mat &samples, icolvec &labels)
void	ecocpak::create_samples_matrix (const vector< ClassData > &v, mat &samples, icolvec &labels, ucolvec &n_per_class)
ClassData	ecocpak::merge_classes_vector (const vector< ClassData > &v)
ClassData	ecocpak::merge_selected_classes (const vector< ClassData > &v, const ucolvec &s)
vector< ClassData >	ecocpak::split_class (const ClassData &c, const ucolvec &indices, const ucolvec &samples_per_cluster, const int are_valid)
ucolvec	ecocpak::complement_sffs (const ucolvec &A, const u32 curSubsetSize)
mat	create_samples_matrix (const vector< ClassData > &v)
void	create_samples_matrix (const vector< ClassData > &v, mat &samples, icolvec &labels)
void	create_samples_matrix (const ClassData &first_class, const ClassData &second_class, mat &samples, icolvec &labels)
void	create_samples_matrix (const vector< ClassData > &v, mat &samples, icolvec &labels, ucolvec &n_per_class)
ClassData	merge_classes_vector (const vector< ClassData > &v)
ClassData	merge_selected_classes (const vector< ClassData > &v, const ucolvec &s)
vector< ClassData >	split_class (const ClassData &c, const ucolvec &indices, const ucolvec &samples_per_cluster, const int are_valid)

---

Function Documentation

ucolvec ecocpak::complement	(	const ucolvec &	A,
		const u32	sizeU
	)

• Function complement returns the complement indices of indices vector A considering as Universal set of indices a linear space of size sizeU.

• Input Arguments:
  • A : Current subset of indices.
  • sizeU : Size of universal set.

• Output Arguments:
  • Void.

• Return Argument:
  • An array of indices which is the complement of A with respect to the universal set U = linspace(0, sizeU - 1, sizeU).

ucolvec ecocpak::complement_sffs	(	const ucolvec &	A,
		const u32	curSubsetSize
	)

• Function complement_sffs() is used to compute the complement labels of first curSubsetSize elements of vector A, with respect to universal set U = linspace(0, sizeU - 1, sizeU) where sizeU is equal to the number of A's elements.

• Auxuliary function for SFFS algorithm.

• Input Arguments:
  • A : An array of integers which its complement we want.
  • curSubsetSize : The size of the array A.

• Output Arguments:
  • Void.

• Return Argument:
  • An array of integers which is the complement of A in respect to the universal set U.

bool ecocpak::contains	(	const ucolvec &	A,
		const u32	n,
		const u32	i
	)

• Function contains() checks whether index i is contained in the first n elements of A.

• Input Arguments:
  • A : Indices vector.
  • n : We examine for membership only the first n elements of A.
  • i : The index we want to check for membership.

• Output Arguments:
  • Void.

• Return Argument:
  • A boolean value which is true if membership of n to A holds, false otherwise.

vector<ClassData> ecocpak::create_class_vector	(	const mat &	samples,
		const icolvec &	labels
	)

• Function create_class_vector() takes a 2D matrix of samples with their respective class labels and creates a vector of classes.

• Input Arguments:
  • samples : An armadillo format 2D matrix which its rows represent the sample vectors.
  • labels : An armadillo format column vector which represents the sample vectors' labels.

• Output Arguments:
  • Void.

• Return Argument:
  • An std vector which contains the created classes as ClassData objects.

void create_samples_matrix	(	const vector< ClassData > &	v,
		mat &	samples,
		icolvec &	labels,
		ucolvec &	n_per_class
	)

• Function create_samples_matrix() creates a samples' matrix with their respective class labels vector from a classes vector.

• This version also outputs a vector of unsigned ints where its positions correspond to the number of samples of each class.

• Overloaded function.

• Input Arguments:
  • v : vector of classes.

• Output Arguments:
  • samples : Samples' matrix.
  • labels : Samples' labels.
  • n_per_class : A vector which holds the number of samples per class.

• Return Argument:
  • Void.

void create_samples_matrix	(	const ClassData &	first_class,
		const ClassData &	second_class,
		mat &	samples,
		icolvec &	labels
	)

• Function create_samples_matrix() creates a samples' matrix with their respective class labels vector from two ClassData objects.

• Overloaded Function.

• Input Arguments:
  • first class : First class.
  • second class : Second class.

• Output Arguments:
  • samples : Samples' matrix.
  • labels : Samples' labels.

• Return Argument:
  • Void.

void create_samples_matrix	(	const vector< ClassData > &	v,
		mat &	samples,
		icolvec &	labels
	)

• Function create_samples_matrix() creates a samples' matrix with their respective class labels vector.

• Overloaded Function.

• Input Arguments:
  • v : vector of classes.

• Output Arguments:
  • samples : Samples' matrix.
  • labels : Samples' labels.

• Return Argument:
  • Void.

mat ecocpak::create_samples_matrix

(

const vector< ClassData > &

v

)

• Function create_samples_matrix() constructs a matrix from a classes vector.

• Overloaded function.

• Input Arguments:
  • v : Vector of classes.

• Output Arguments:
  • Void.

• Returns Argument:
  • Constructed matrix.

void ecocpak::create_samples_matrix	(	const vector< ClassData > &	v,
		mat &	samples,
		icolvec &	labels
	)

• Function create_samples_matrix() creates a samples' matrix with their respective class labels vector.

• Overloaded Function.

• Input Arguments:
  • v : vector of classes.

• Output Arguments:
  • samples : Samples' matrix.
  • labels : Samples' labels.

• Return Argument:
  • Void.

void ecocpak::create_samples_matrix	(	const ClassData &	first_class,
		const ClassData &	second_class,
		mat &	samples,
		icolvec &	labels
	)

• Function create_samples_matrix() creates a samples' matrix with their respective class labels vector from two ClassData objects.

• Overloaded Function.

• Input Arguments:
  • first class : First class.
  • second class : Second class.

• Output Arguments:
  • samples : Samples' matrix.
  • labels : Samples' labels.

• Return Argument:
  • Void.

mat create_samples_matrix

(

const vector< ClassData > &

v

)

• Function create_samples_matrix() constructs a matrix from a classes vector.

• Overloaded function.

• Input Arguments:
  • v : Vector of classes.

• Output Arguments:
  • Void.

• Returns Argument:
  • Constructed matrix.

void ecocpak::create_samples_matrix	(	const vector< ClassData > &	v,
		mat &	samples,
		icolvec &	labels,
		ucolvec &	n_per_class
	)

• Function create_samples_matrix() creates a samples' matrix with their respective class labels vector from a classes vector.

• This version also outputs a vector of unsigned ints where its positions correspond to the number of samples of each class.

• Overloaded function.

• Input Arguments:
  • v : vector of classes.

• Output Arguments:
  • samples : Samples' matrix.
  • labels : Samples' labels.
  • n_per_class : A vector which holds the number of samples per class.

• Return Argument:
  • Void.

ClassData ecocpak::merge_classes_vector

(

const vector< ClassData > &

v

)

• Function merge_classes_vector() merges the classes of its input class vector into one single ClassData object (i.e., class).

• Input Arguments:
  • v : vector of classes.

• Output Arguments:
  • Void.

• Return Argument:
  • ClassData object -- representing the merged product class.

ClassData merge_classes_vector

(

const vector< ClassData > &

v

)

• Function merge_classes_vector() merges the classes of its input class vector into one single ClassData object (i.e., class).

• Input Arguments:
  • v : vector of classes.

• Output Arguments:
  • Void.

• Return Argument:
  • ClassData object -- representing the merged product class.

ClassData merge_selected_classes	(	const vector< ClassData > &	v,
		const ucolvec &	s
	)

• Function merge_selected_classes() merges a subset of classes from an input vector of classes v.

• The selected classes' indices are supplied in the s input vector.

• Input arguments:
  • v : Vector of classes.
  • s : Array which holds the labels of merging classes.

• Output Arguments:
  • Void.

• Return Argument:
  • ClassData object

ClassData ecocpak::merge_selected_classes	(	const vector< ClassData > &	v,
		const ucolvec &	s
	)

• Function merge_selected_classes() merges a subset of classes from an input vector of classes v.

• The selected classes' indices are supplied in the s input vector.

• Input arguments:
  • v : Vector of classes.
  • s : Array which holds the labels of merging classes.

• Output Arguments:
  • Void.

• Return Argument:
  • ClassData object

vector<ClassData> split_class	(	const ClassData &	c,
		const ucolvec &	indices,
		const ucolvec &	samples_per_cluster,
		const int	are_valid
	)

• Function split_class() splits a ClassData object into classes according to supplied indices vector.

• Current version supports split only into two classes.

• Input Arguments:
  • c : Class to be split.
  • indices : Indices of split.
  • samples_per_cluster : Number of samples in each cluster.
  • are_valid : Indicator which signifies whether the new classes are valid or they are intermidiate products.

• Output Arguments:
  • Void.

• Return Argument:
  • A vector which holds the newly created classes.

vector<ClassData> ecocpak::split_class	(	const ClassData &	c,
		const ucolvec &	indices,
		const ucolvec &	samples_per_cluster,
		const int	are_valid
	)

• Function split_class() splits a ClassData object into classes according to supplied indices vector.

• Current version supports split only into two classes.

• Input Arguments:
  • c : Class to be split.
  • indices : Indices of split.
  • samples_per_cluster : Number of samples in each cluster.
  • are_valid : Indicator which signifies whether the new classes are valid or they are intermidiate products.

• Output Arguments:
  • Void.

• Return Argument:
  • A vector which holds the newly created classes.