Collaboration diagram for Overloaded update methods:

Modules
	Improper prior methods

	0-mean prior methods

	non-0-mean prior methods

Functions
void	RanIndex::update (const Grp &theta, const Grp &mu, const vector< SigmaI > &SigI, const MixP &p)
	Update mixture model with multiple covariances. More...

void	RanIndex::update (const Grp &theta, const Grp &mu, const SigmaI &SigI, const MixP &p)
	Update mixture model with a single covariance. More...

virtual void	RanIndex::update (const Grp &y, const SigmaI &SigIe, BetaGrpBVSR *theta, const SigmaI &SigIp)
	Variable selection update. More...

void	Apex::update (const Grp &y, const gsl_matrix *xi, const SigmaI &SigIm)
	Unreplicated Gaussian update. More...

void	Apex::update (const Grp &y, const gsl_matrix *xi, const SigmaI &SigIm, const RanIndex &ind)
	Replicated Gaussian update. More...

void	Apex::update (const Grp &y, const gsl_matrix *xi, const Qgrp &q, const SigmaI &SigIm, const RanIndex &ind)
	Replicated Student- \(t\) update. More...

virtual void	MuGrpMiss::update (const Grp &mu, const SigmaI &SigIm)
	Standard Gaussian imputation. More...

virtual void	MuGrpMiss::update (const Grp &mu, const SigmaI &SigIm, const SigmaI &SigIp)
	Gaussian imputation with a prior. More...

virtual void	MuGrpEE::update (const Grp &muPr, const SigmaI &SigIm)
	Gaussian prior. More...

virtual void	MuGrpEE::update (const Grp &muPr, const Qgrp &q, const SigmaI &SigIm)
	Student- \(t\) prior. More...

void	BetaGrpSnp::update (const Grp &dat, const SigmaI &SigIm)
	Response update function. More...

void	BetaGrpSnpMiss::update (const Grp &dat, const SigmaI &SigIm)
	Response update function. More...

virtual void	SigmaI::update (const Grp &dat)
	Basic Gaussian update. More...

virtual void	SigmaI::update (const Grp &dat, const Grp &mu)
	Gaussian update with a mean. More...

virtual void	SigmaI::update (const Grp &dat, const Qgrp &q)
	Basic Student- \(t\) update. More...

virtual void	SigmaI::update (const Grp &dat, const Grp &mu, const Qgrp &q)
	Student- \(t\) update with a mean. More...

virtual void	Qgrp::update (const Grp &dat, const Grp &mu, const SigmaI &SigI)
	Update with a mean. More...

virtual void	Qgrp::update (const Grp &dat, const SigmaI &SigI)
	Basic update. More...

void	MixP::update (const RanIndex &Nvec)
	Gibbs update function. More...

Detailed Description

Update functions generate new stochastic values of a given parameter or set of parameters as we step through the Markov chain iteration. These are mostly Gibbs updates, but occasional Metropolis steps are used in special cases.

Function Documentation

◆ update() [1/19]

void SigmaI::update ( const Grp & dat )

virtual

Basic Gaussian update.

The data are assumed already centered, so the update is based on the simple cross-product of the data.

Parameters

[in] Grp& data

Reimplemented in SigmaIblk.

◆ update() [2/19]

void SigmaI::update	(	const Grp &	dat,
		const Grp &	mu
	)

virtual

Gaussian update with a mean.

The mean value is subtracted from the data according to the up-pointing RanIndex in the data object.

Parameters

[in]	Grp&	data
[in]	Grp&	mean

Reimplemented in SigmaIblk.

◆ update() [3/19]

void SigmaI::update	(	const Grp &	dat,
		const Grp &	mu,
		const Qgrp &	q
	)

virtual

Student- \(t\) update with a mean.

The mean value is subtracted from the data according to the up-pointing RanIndex in the data object.

Parameters

[in]	Grp&	data
[in]	Grp&	mean
[in]	Qgrp&	scale parameter

Reimplemented in SigmaIpex.

◆ update() [4/19]

void Qgrp::update	(	const Grp &	dat,
		const Grp &	mu,
		const SigmaI &	SigI
	)

virtual

Update with a mean.

The mean value is subtracted from the data according to the up-pointing RanIndex in the data object.

Parameters

[in]	Grp&	data
[in]	Grp&	mean
[in]	SigmaI&	inverse-covariance

Reimplemented in QgrpPEX.

◆ update() [5/19]

void SigmaI::update	(	const Grp &	dat,
		const Qgrp &	q
	)

virtual

Basic Student- \(t\) update.

The data are assumed already centered, so the update is based on the simple cross-product of the data and the current value of the scale parameter.

Parameters

[in]	Grp&	data
[in]	Qgrp&	scale parameter

Reimplemented in SigmaIpex.

◆ update() [6/19]

void Qgrp::update	(	const Grp &	dat,
		const SigmaI &	SigI
	)

virtual

Basic update.

The data are assumed already centered, so the update is based on the simple cross-product of the data and the current value of the inverse-covariance.

Parameters

[in]	Grp&	data
[in]	SigmaI&	inverse-covariance

Reimplemented in QgrpPEX.

◆ update() [7/19]

void BetaGrpSnp::update	(	const Grp &	dat,
		const SigmaI &	SigIm
	)

virtual

Response update function.

Unlike standard update functions, this one only saves MCMC samples of the response. The covariance provided is ignored.
The actual regression is performed on the point estimates of response values calculated as means of the stored MCMC values, and is done at the end by envoking the dump() function.

Parameters

[in]	Grp&	response variable to be saved
[in]	SigmaI&	inverse-covariance (ignored)

Reimplemented from MuGrp.

◆ update() [8/19]

void BetaGrpSnpMiss::update	(	const Grp &	dat,
		const SigmaI &	SigIm
	)

virtual

Response update function.

Unlike standard update functions, this one only saves MCMC samples of the response. The covariance provided is ignored. The actual regression is performed on the point estimates of response values calculated as means of the stored MCMC values, and is done at the end by envoking the dump() function.

Parameters

[in]	Grp&	response variable to be saved
[in]	SigmaI&	inverse-covariance (ignored)

Reimplemented from MuGrp.

◆ update() [9/19]

void MuGrpMiss::update	(	const Grp &	mu,
		const SigmaI &	SigIm
	)

virtual

Standard Gaussian imputation.

If some data are present, performs standard Gaussian marginal imputation (desribed in, e.g., [chatfield80] ) with the provided Grp object as a mean and the SigmaI object as the inverse-covariance. If no data for a row are present, simply replaces the row values by a Gaussian sample with mean and inverse-covariance provided. Rows with no missing data are ignored.

Parameters

[in]	Grp&	mean
[in]	SigmaI&	inverse-covariance

Reimplemented from MuGrp.

Reimplemented in MuGrpEEmiss.

◆ update() [10/19]

void MuGrpMiss::update	(	const Grp &	mu,
		const SigmaI &	SigIm,
		const SigmaI &	SigIp
	)

virtual

Gaussian imputation with a prior.

If some data are present, performs Gaussian marginal imputation (desribed in, e.g., [chatfield80] ) with the provided Grp object as a mean and the SigmaI object as the inverse-covariance, but with a 0-mean prior. If no data for a row are present, simply replaces the row values by a Gaussian sample with mean and inverse-covariance provided. Rows with no missing data are ignored.

Warning: Has not been extensively tested

Parameters

[in]	Grp&	mean
[in]	SigmaI&	inverse-covariance
[in]	SigmaI&	prior inverse-covariance

Reimplemented from MuGrp.

◆ update() [11/19]

void MuGrpEE::update	(	const Grp &	muPr,
		const Qgrp &	q,
		const SigmaI &	SigIm
	)

virtual

Student- \(t\) prior.

The Grp object contains the prior means and the SigmaI object – the prior inverse-covariance for the sampling of data values. The upper index of the object must have the same number of groups as the number of rows in the prior matrix addressed by fMat(). While the sampling is independent, the prior inverse variances for each trait are taken from the diagonal of the inverse-covariance matrix (in the SigmaI object), and are thus influenced by any correlated traits.

Parameters

[in]	Grp&	prior mean
[in]	Qgrp&	Student- \(t\) weights
[in]	SigmaI&	prior inverse-covariance

Reimplemented from MuGrp.

◆ update() [12/19]

void MuGrpEE::update	(	const Grp &	muPr,
		const SigmaI &	SigIm
	)

virtual

Gaussian prior.

The Grp object contains the prior means and the SigmaI object – the prior inverse-covariance for the sampling of data values. The upper index of the object must have the same number of groups as the number of rows in the prior matrix addressed by fMat(). While the sampling is independent, the prior inverse variances for each trait are taken from the diagonal of the inverse-covariance matrix (in the SigmaI object), and are thus influenced by any correlated traits.

Parameters

[in]	Grp&	prior mean
[in]	SigmaI&	prior inverse-covariance

Reimplemented from MuGrp.

◆ update() [13/19]

void RanIndex::update	(	const Grp &	theta,
		const Grp &	mu,
		const SigmaI &	SigI,
		const MixP &	p
	)

Update mixture model with a single covariance.

Updates the group relationships for Gaussian mixture models when each group has a different mean but all covrariances are the same.

Parameters

[in]	Grp&	data
[in]	Grp&	group means
[in]	SigmaI&	common inverse-covariance
[in]	MixP&	prior proportions

◆ update() [14/19]

void RanIndex::update	(	const Grp &	theta,
		const Grp &	mu,
		const vector< SigmaI > &	SigI,
		const MixP &	p
	)

Update mixture model with multiple covariances.

Updates the group relationships for Gaussian mixture models when each group has a different mean and covariance.

Parameters

[in]	Grp&	data
[in]	Grp&	group means
[in]	vector<SigmaI>&	vector of inverse-covariances
[in]	MixP&	prior proportions

◆ update() [15/19]

void Apex::update	(	const Grp &	y,
		const gsl_matrix *	xi,
		const Qgrp &	q,
		const SigmaI &	SigIm,
		const RanIndex &	ind
	)

Replicated Student- \(t\) update.

Student- \(t\) data with replication, i.e., the number of rows in \( \boldsymbol{Y} \) is larger than the number of rows in \( \boldsymbol{\Xi} \).

Parameters

[in]	Grp&	data
[in]	gsl_matrix*	"raw" location parameter matrix
[in]	Qgrp&	Student- \(t\) weights
[in]	SigmaI&	data inverse-covariance matrix
[in]	RanIndex&	index relating data rows to rows in \( \boldsymbol{\Xi} \)

◆ update() [16/19]

void Apex::update	(	const Grp &	y,
		const gsl_matrix *	xi,
		const SigmaI &	SigIm
	)

Unreplicated Gaussian update.

Gaussian data with no replication.

Parameters

[in]	Grp&	data
[in]	gsl_matrix*	"raw" location parameter matrix
[in]	SigmaI&	data inverse-covariance matrix

◆ update() [17/19]

void Apex::update	(	const Grp &	y,
		const gsl_matrix *	xi,
		const SigmaI &	SigIm,
		const RanIndex &	ind
	)

Replicated Gaussian update.

Gaussian data with replication, i.e., the number of rows in \( \boldsymbol{Y} \) is larger than the number of rows in \( \boldsymbol{\Xi} \).

Parameters

[in]	Grp&	data
[in]	gsl_matrix*	"raw" location parameter matrix
[in]	SigmaI&	data inverse-covariance matrix
[in]	RanIndex&	index relating data rows to rows in \( \boldsymbol{\Xi} \)

◆ update() [18/19]

virtual void RanIndex::update	(	const Grp &	y,
		const SigmaI &	SigIe,
		BetaGrpBVSR *	theta,
		const SigmaI &	SigIp
	)

inlinevirtual

Variable selection update.

Update for a mixture that includes point-mass at 0. Only implemented in the derived class.

Parameters

[in]	Grp&	data
[in]	SigmaI&	likelihood inverse-covariance
[in]	BetaGrpBVSR*	location parameter values
[in]	SigmaI&	prior inverse-covariance

Reimplemented in RanIndexVS.

◆ update() [19/19]

void MixP::update ( const RanIndex & Nvec )

Gibbs update function.

Performs standard Gibbs mixture updating (e.g., [gelman04] ) with samples from a Dirichlet distribution.

Parameters

[in] RanIndex& data index of which elements belong to which category

Modules

Functions

Detailed Description

Function Documentation

◆ update() [1/19]

◆ update() [2/19]

◆ update() [3/19]

◆ update() [4/19]

◆ update() [5/19]

◆ update() [6/19]

◆ update() [7/19]

◆ update() [8/19]

◆ update() [9/19]

◆ update() [10/19]

◆ update() [11/19]

◆ update() [12/19]

◆ update() [13/19]

◆ update() [14/19]

◆ update() [15/19]

◆ update() [16/19]

◆ update() [17/19]

◆ update() [18/19]

◆ update() [19/19]