tatami_stats::variances Namespace Reference

Functions for computing dimension-wise variances. More...

Classes
struct	Options
	Variance calculation options. More...
class	RunningDense
	Running variances from dense data. More...
class	RunningSparse
	Running variances from sparse data. More...

Functions
template<typename Output_ = double, typename Value_ , typename Index_ >
std::pair< Output_, Output_ >	direct (const Value_ *value, Index_ num_nonzero, Index_ num_all, bool skip_nan)
template<typename Output_ = double, typename Value_ , typename Index_ >
std::pair< Output_, Output_ >	direct (const Value_ *ptr, Index_ num, bool skip_nan)
template<typename Value_ , typename Index_ , typename Output_ >
void	apply (bool row, const tatami::Matrix< Value_, Index_ > &mat, Output_ *output, const Options &vopt)
template<typename Output_ = double, typename Value_ , typename Index_ >
std::vector< Output_ >	by_column (const tatami::Matrix< Value_, Index_ > &mat, const Options &vopt)
template<typename Output_ = double, typename Value_ , typename Index_ >
std::vector< Output_ >	by_row (const tatami::Matrix< Value_, Index_ > &mat, const Options &vopt)

Detailed Description

Functions for computing dimension-wise variances.

Function Documentation

direct() [1/2]

template<typename Output_ = double, typename Value_ , typename Index_ >

std::pair< Output_, Output_ > tatami_stats::variances::direct	(	const Value_ *	value,
		Index_	num_nonzero,
		Index_	num_all,
		bool	skip_nan )

Compute the mean and variance from a sparse objective vector. This uses the standard two-pass algorithm with naive accumulation of the sum of squared differences; thus, it is best used with a sufficiently high-precision Output_ like double.

Template Parameters

Output_	Floating-point type of the output data. This should be capable of storing NaNs.
Value_	Numeric type of the input data.
Index_	Integer type of the row/column indices.

Parameters

[in]	value	Pointer to an array of length num, containing the values of the structural non-zeros.
	num_nonzero	Length of the array pointed to by value.
	num_all	Length of the objective vector, including the structural zeros not in value. This should be greater than or equal to num_nonzero.
	skip_nan	See Options::skip_nan.

Returns

The sample mean and variance of values in the sparse array. This may be NaN if there are not enough (non-NaN) values in value.

direct() [2/2]

template<typename Output_ = double, typename Value_ , typename Index_ >

std::pair< Output_, Output_ > tatami_stats::variances::direct	(	const Value_ *	ptr,
		Index_	num,
		bool	skip_nan )

Compute the mean and variance from a dense objective vector. This uses the standard two-pass algorithm with naive accumulation of the sum of squared differences; thus, it is best used with a sufficiently high-precision Output_ like double.

Template Parameters

Output_	Floating-point type of the output data. This should be capable of storing NaNs.
Value_	Numeric type of the input data.
Index_	Integer type of the row/column indices.

Parameters

[in]	ptr	Pointer to an array of length num, containing the values of the objective vector.
	num	Length of the objective vector, i.e., length of the array at ptr.
	skip_nan	See Options::skip_nan.

Returns

The sample mean and variance of values in [ptr, ptr + num). This may be NaN if there are not enough (non-NaN) values in ptr.

apply()

template<typename Value_ , typename Index_ , typename Output_ >

void tatami_stats::variances::apply	(	bool	row,
		const tatami::Matrix< Value_, Index_ > &	mat,
		Output_ *	output,
		const Options &	vopt )

Compute variances for each element of a chosen dimension of a tatami::Matrix. This may use either Welford's method or the standard two-pass method, depending on the dimension in row and the preferred access dimension of p.

Template Parameters

Value_	Numeric type of the input data.
Index_	Integer type of the row/column indices.
Output_	Floating-point type of the output data. This should be capable of storing NaNs.

Parameters

	row	Whether to compute the variance for each row. If false, the variance is computed for each column instead.
	mat	Instance of a tatami::Matrix.
[out]	output	Pointer to an array of length equal to the number of rows (if row = true) or columns (otherwise). On output, this will contain the row/column variances.
	vopt	Variance calculation options.

by_column()

template<typename Output_ = double, typename Value_ , typename Index_ >

std::vector< Output_ > tatami_stats::variances::by_column	(	const tatami::Matrix< Value_, Index_ > &	mat,
		const Options &	vopt )

Wrapper around apply() for column variances.

Template Parameters

Output_	Floating-point type of the output data. This should be capable of storing NaNs.
Value_	Numeric type of the input data.
Index_	Integer type of the row/column indices.

Parameters

mat	Instance of a tatami::Matrix.
vopt	Variance calculation options.

Returns

A vector of length equal to the number of columns, containing the column variances.

by_row()

template<typename Output_ = double, typename Value_ , typename Index_ >

std::vector< Output_ > tatami_stats::variances::by_row	(	const tatami::Matrix< Value_, Index_ > &	mat,
		const Options &	vopt )

Wrapper around apply() for column variances.

Template Parameters

Output_	Floating-point type of the output data. This should be capable of storing NaNs.
Value_	Numeric type of the input data.
Index_	Integer type of the row/column indices.

Parameters

mat	Instance of a tatami::Matrix.
vopt	Variance calculation options.

Returns

A vector of length equal to the number of rows, containing the row variances.