STSLIB::StatsMcCorrection Class Reference

Bonferroni, Holm-Bonferroni and Benjamini-Hochberg FDR adjustments for mass-univariate p-value maps. More...

#include <sts_correction.h>

Public Types
enum class	FdrMethod { Independent , NegativelyCorrelated }
using	RejectMask = Eigen::Array<bool, Eigen::Dynamic, Eigen::Dynamic>

Static Public Member Functions
static Eigen::MatrixXd	bonferroni (const Eigen::MatrixXd &pValues, double alpha=0.05, RejectMask *reject=nullptr)
static Eigen::MatrixXd	holmBonferroni (const Eigen::MatrixXd &pValues)
static Eigen::MatrixXd	fdr (const Eigen::MatrixXd &pValues, double alpha=0.05, FdrMethod method=FdrMethod::Independent, RejectMask *reject=nullptr)

Detailed Description

Bonferroni, Holm-Bonferroni and Benjamini-Hochberg FDR adjustments for mass-univariate p-value maps.

Multiple comparison correction methods.

Definition at line 60 of file sts_correction.h.

Member Typedef Documentation

RejectMask

using STSLIB::StatsMcCorrection::RejectMask = Eigen::Array<bool, Eigen::Dynamic, Eigen::Dynamic>

Hypotheses rejected by a correction, same shape as the p-values.

Definition at line 67 of file sts_correction.h.

Member Enumeration Documentation

FdrMethod

enum class STSLIB::StatsMcCorrection::FdrMethod

strong

Dependence assumed by the FDR correction (mne's method "indep" / "negcorr").

Enumerator
Independent	Benjamini-Hochberg: independent or positively correlated tests.
NegativelyCorrelated	Benjamini-Yekutieli: arbitrary dependence.

Definition at line 72 of file sts_correction.h.

Member Function Documentation

bonferroni()

MatrixXd StatsMcCorrection::bonferroni ( const Eigen::MatrixXd & pValues, double alpha = 0.05, RejectMask * reject = nullptr )

static

Bonferroni correction as mne.stats.bonferroni_correction: corrected_p = min(p * n, 1.0).

Parameters

[in]	pValues	Matrix of p-values.
[in]	alpha	Significance level for reject.
[out]	reject	If not null, receives corrected_p < alpha.

Returns

Corrected p-values (same shape as input).

Definition at line 43 of file sts_correction.cpp.

fdr()

MatrixXd StatsMcCorrection::fdr ( const Eigen::MatrixXd & pValues, double alpha = 0.05, FdrMethod method = FdrMethod::Independent, RejectMask * reject = nullptr )

static

False Discovery Rate correction as mne.stats.fdr_correction.

Parameters

[in]	pValues	Matrix of p-values.
[in]	alpha	False discovery rate for reject.
[in]	method	Benjamini-Hochberg or Benjamini-Yekutieli.
[out]	reject	If not null, receives the hypotheses rejected by the step-up procedure.

Returns

Corrected p-values (same shape as input).

Definition at line 88 of file sts_correction.cpp.

holmBonferroni()

MatrixXd StatsMcCorrection::holmBonferroni ( const Eigen::MatrixXd & pValues )

static

Holm-Bonferroni step-down correction.

Parameters

[in]

pValues

Matrix of p-values.

Returns

Corrected p-values (same shape as input).

Definition at line 53 of file sts_correction.cpp.

---

The documentation for this class was generated from the following files:

• src/libraries/sts/sts_correction.h
• src/libraries/sts/sts_correction.cpp