sas_core.hpp File Reference

Core numerical utilities used across the project. More...

#include <eigen3/Eigen/Dense>

Go to the source code of this file.

Enumerations
enum class	sas::Statistics { Mean }
	Statistical measures supported by sas_core utilities. More...

Functions
template<class T>
constexpr T	sas::incremental_mean (const T &current_mean, const int &current_number_of_samples, const T &new_sample)
	incremental_mean a simple implementation of incremental mean, for the many cases in which keeping a vector of all values would be impractical.
VectorXd	sas::concatenate (const VectorXd &a, const VectorXd &b)
	Concatenate two vectors by appending b after a.
VectorXd	sas::concatenate (const std::vector< VectorXd > &as)
	Concatenate a list of vectors into a single vector.
MatrixXd	sas::vstack (const MatrixXd &A, const MatrixXd &B)
	Stack two matrices vertically (A above B).
MatrixXd	sas::block_diag (const std::vector< MatrixXd > &As)
	Create a block-diagonal matrix from a list of matrices.
std::vector< VectorXd >	sas::split (const VectorXd &a, const std::vector< int > &ns)
	Split a vector into pieces with sizes specified by ns.

Detailed Description

Core numerical utilities used across the project.

This header contains lightweight, frequently used functions such as incremental mean computation, vector/matrix concatenation, block-diagonal assembly and vector splitting utilities. These utilities are intended to be header-only and efficient for use in real-time control loops.

Enumeration Type Documentation

Statistics

enum class sas::Statistics

strong

Statistical measures supported by sas_core utilities.

Enumeration of lightweight statistics that classes and functions may compute (for example, Clock statistics). Add new entries here as new statistics are supported.

Function Documentation

block_diag()

MatrixXd sas::block_diag

(

const std::vector< MatrixXd > &

As

)

Create a block-diagonal matrix from a list of matrices.

block_diag creates a block diagonal matrix using an input of std::vector<MatrixXd>. e.g. if As= [A, B, C], then block_diag(As) = |A 0 0| |0 B 0| |0 0 C|

Parameters

As	Vector of matrices to place on the block diagonal.

Returns

Block-diagonal matrix containing the input matrices along its diagonal.

Parameters

As	the std::vector<MatrixXd> contaning the matrix to form the block diagonal matrix.

Returns

the block diagonal matrix.

concatenate() [1/2]

VectorXd sas::concatenate

(

const std::vector< VectorXd > &

as

)

Concatenate a list of vectors into a single vector.

concatenate a std::vector of VectorXd.

Parameters

as	Vector of vectors to concatenate in order.

Returns

Concatenated vector containing the elements of each input vector in order.

Parameters

a	an std::vector of VectorXd.

Returns

the result of the concatenated vectors.

concatenate() [2/2]

VectorXd sas::concatenate	(	const VectorXd &	a,
		const VectorXd &	b )

Concatenate two vectors by appending b after a.

concatenate two VectorXd.

Parameters

a	First vector.
b	Second vector.

Returns

Concatenated vector containing all elements of a followed by b.

Parameters

a	a VectorXd.
b	a VectorXd.

Returns

the result of the concatenated vectors.

incremental_mean()

template<class T>

T sas::incremental_mean ( const T & current_mean, const int & current_number_of_samples, const T & new_sample )

constexpr

incremental_mean a simple implementation of incremental mean, for the many cases in which keeping a vector of all values would be impractical.

Parameters

current_mean	the current value of the mean.
current_number_of_samples	the current number of samples, not considering the new_sample.
new_sample	the new sample that will change the mean.

Exceptions

std::range_error

if current_number_of_samples is negative.

Returns

the incremental mean, considering the new_sample.

split()

std::vector< VectorXd > sas::split	(	const VectorXd &	a,
		const std::vector< int > &	ns )

Split a vector into pieces with sizes specified by ns.

split splits the input VectorXd as into a set of subvectors defined by ns.

Parameters

a	Vector to split.
ns	Sizes of each piece; their sum must equal a.size().

Returns

Vector containing the split VectorXd pieces.

Parameters

as	the VectorXd to be split.
ns	the sizes of the subvectors.

Returns

an std::vector<VectorXd> of the splitted vectors.

vstack()

MatrixXd sas::vstack	(	const MatrixXd &	A,
		const MatrixXd &	B )

Stack two matrices vertically (A above B).

vstack vertically (row-wise) stack two MatrixXd.

Parameters

A	Top matrix.
B	Bottom matrix.

Returns

Matrix formed by stacking A on top of B. Columns must match.

Exceptions

std::range_error

if A and B have different numbers of columns.

Parameters

A	the first MatrixXd.
B	the second MatrixXd.

Returns

the vstacked MatrixXd.

Exceptions

a	std::range_error if A and B don't have the same number of columns.

Note

returns an empty matrix if both arguments are empty or return the other argument of only one of the arguments is empty.