Mymath Library

Math and matrix utilities. More...

Functions
void	transpone (double *arr, unsigned int Nrows, unsigned int Ncols)
	Transposes a matrix in-place. More...
void	invert (double *arr, unsigned int Nrows, unsigned int Ncols)
	Inverts a square matrix in-place. More...
double *	mulmat (const double *matrix0, unsigned int Nrows, unsigned int Ncols, const double *matrix1, unsigned int Mrows, unsigned int Mcols)
	Multiplies two matrices and returns the result as a new matrix. More...
void	matlinreg (double coeffs[2], const double *xmat, unsigned int Nrows, unsigned int Ncols, double *vec, const double *weight)
	Performs linear regression on a dataset with optional weighting. More...
void	fitNpeaks (double *intercept, double *slope, unsigned int curPos, const struct myarr *maxvals, const struct myarr *maxes, const struct myarr *subpos, unsigned int npeaks, double SDthreshold)
	Fits a line to N peaks in the data, with outlier rejection. More...
void	fastlinreg (double coeffs[2], const double *xmat, unsigned int Npoints, const double *vec, const double *weightArr)
	Performs fast weighted linear regression. More...
size_t	getmaxpos (const int *array, size_t ArrayLength)
	Returns the index of the maximum value in an integer array. More...
void	linreg (const double *xarr, const double *yarr, size_t ArrayLength, double *intercept, double *slope, double *stdev)
	Performs simple linear regression on two arrays. More...
int	shiftHalf (size_t value, size_t ArrayLength)
	Shifts a value by half the array length, wrapping around. More...

Detailed Description

Math and matrix utilities.

Function Documentation

fastlinreg()

void fastlinreg	(	double	coeffs[2],
		const double *	xmat,
		unsigned int	Npoints,
		const double *	vec,
		const double *	weightArr
	)

Performs fast weighted linear regression.

Parameters

coeffs	Array to store the resulting coefficients (intercept and slope).
xmat	The array of x values.
Npoints	The number of points.
vec	The array of y values.
weightArr	The array of weights.

fitNpeaks()

void fitNpeaks	(	double *	intercept,
		double *	slope,
		unsigned int	curPos,
		const struct myarr *	maxvals,
		const struct myarr *	maxes,
		const struct myarr *	subpos,
		unsigned int	npeaks,
		double	SDthreshold
	)

Fits a line to N peaks in the data, with outlier rejection.

Parameters

intercept	Pointer to store the intercept.
slope	Pointer to store the slope.
curPos	The current position in the data.
maxvals	Array of weights.
maxes	Array of y values.
subpos	Array of sub-positions.
npeaks	Number of peaks to fit.
SDthreshold	Standard deviation threshold for outlier rejection.

getmaxpos()

size_t getmaxpos	(	const int *	array,
		size_t	ArrayLength
	)

Returns the index of the maximum value in an integer array.

Parameters

array	The array to search.
ArrayLength	The number of elements in the array.

Returns

The index of the maximum value.

invert()

void invert	(	double *	arr,
		unsigned int	Nrows,
		unsigned int	Ncols
	)

Inverts a square matrix in-place.

Parameters

arr	The square matrix to invert, stored as a 1D array in row-major order. The result will overwrite the original matrix.
Nrows	The number of rows (and columns) in the square matrix.
Ncols	The number of columns (and rows) in the square matrix.

linreg()

void linreg	(	const double *	xarr,
		const double *	yarr,
		size_t	ArrayLength,
		double *	intercept,
		double *	slope,
		double *	stdev
	)

Performs simple linear regression on two arrays.

Parameters

xarr	The array of x values.
yarr	The array of y values.
ArrayLength	The number of elements in the arrays.
intercept	Pointer to store the intercept.
slope	Pointer to store the slope.
stdev	Pointer to store the standard deviation of the fit.

matlinreg()

void matlinreg	(	double	coeffs[2],
		const double *	xmat,
		unsigned int	Nrows,
		unsigned int	Ncols,
		double *	vec,
		const double *	weight
	)

Performs linear regression on a dataset with optional weighting.

Parameters

coeffs	An array of size 2 where the resulting coefficients (intercept and slope) will be stored.
xmat	The design matrix, stored as a 1D array in row-major order.
Nrows	The number of rows in the design matrix.
Ncols	The number of columns in the design matrix.
vec	The response vector, stored as a 1D array.
weight	An optional array of weights for each data point, stored as a 1D array. If NULL, unweighted regression is performed.

mulmat()

double * mulmat	(	const double *	matrix0,
		unsigned int	Nrows,
		unsigned int	Ncols,
		const double *	matrix1,
		unsigned int	Mrows,
		unsigned int	Mcols
	)

Multiplies two matrices and returns the result as a new matrix.

Parameters

matrix0	The first matrix, stored as a 1D array in row-major order.
Nrows	The number of rows in the first matrix.
Ncols	The number of columns in the first matrix.
matrix1	The second matrix, stored as a 1D array in row-major order.
Mrows	The number of rows in the second matrix.
Mcols	The number of columns in the second matrix.

Returns

A pointer to the resulting matrix, stored as a 1D array in row-major order. The caller is responsible for freeing this memory.

shiftHalf()

int shiftHalf	(	size_t	value,
		size_t	ArrayLength
	)

Shifts a value by half the array length, wrapping around.

Parameters

value	The value to shift.
ArrayLength	The length of the array.

Returns

The shifted value, wrapped around the array length.

transpone()

void transpone	(	double *	arr,
		unsigned int	Nrows,
		unsigned int	Ncols
	)

Transposes a matrix in-place.

Parameters

arr	The matrix to transpose, stored as a 1D array in row-major order.
Nrows	The number of rows in the original matrix.
Ncols	The number of columns in the original matrix.