Class MatrixVectorProductOp

Provides outgoing messages for the following factors:

, given random arguments to the function.

Inheritance

Object

MatrixVectorProductOp

Inherited Members

Object.Equals(Object)

Object.Equals(Object, Object)

Object.GetHashCode()

Object.GetType()

Object.MemberwiseClone()

Object.ReferenceEquals(Object, Object)

Object.ToString()

Namespace: Microsoft.ML.Probabilistic.Factors

Assembly: Microsoft.ML.Probabilistic.dll

Syntax

[FactorMethod(typeof(Factor), "Product", new Type[]{typeof(Matrix), typeof(Vector)})]
[FactorMethod(typeof(Factor), "Product", new Type[]{typeof(double[, ]), typeof(Vector)})]
[Buffers(new string[]{"BMean", "BVariance"})]
[Quality(QualityBand.Stable)]
public static class MatrixVectorProductOp

Fields

UseAccurateMethod

Declaration

public static bool UseAccurateMethod

Field Value

Type	Description
Boolean

Methods

AAverageConditional(VectorGaussian, DistributionArray2D<Gaussian, Double>, Vector, PositiveDefiniteMatrix, DistributionStructArray2D<Gaussian, Double>)

EP message to a.

Declaration

public static DistributionStructArray2D<Gaussian, double> AAverageConditional(VectorGaussian product, DistributionArray2D<Gaussian, double> A, Vector BMean, PositiveDefiniteMatrix BVariance, DistributionStructArray2D<Gaussian, double> result)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
DistributionArray2D<Gaussian, Double>	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
DistributionStructArray2D<Gaussian, Double>	result	Modified to contain the outgoing message.

Returns

Type	Description
DistributionStructArray2D<Gaussian, Double>	`result`

Remarks

The outgoing message is a distribution matching the moments of a as the random arguments are varied. The formula is proj[p(a) sum_(product) p(product) factor(product,a,b)]/p(a).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

AverageLogFactor(VectorGaussian, Matrix, VectorGaussian)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(VectorGaussian product, Matrix A, VectorGaussian B)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	B	Incoming message from `b`.

Returns

Type	Description
Double	Zero.

Remarks

In Variational Message Passing, the evidence contribution of a deterministic factor is zero. Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

AverageLogFactor(Vector, Matrix, VectorGaussian, Vector, PositiveDefiniteMatrix)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(Vector product, Matrix A, VectorGaussian B, Vector BMean, PositiveDefiniteMatrix BVariance)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	B	Incoming message from `b`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.

Returns

Type	Description
Double	Zero.

Remarks

In Variational Message Passing, the evidence contribution of a deterministic factor is zero. Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

AverageLogFactor(Vector, Matrix, Vector)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(Vector product, Matrix A, Vector B)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
Vector	B	Constant value for `b`.

Returns

Type	Description
Double	Zero.

Remarks

The formula for the result is log(factor(product,a,b)). Adding up these values across all factors and variables gives the log-evidence estimate for VMP.

BAverageConditional(VectorGaussian, DistributionArray2D<Gaussian, Double>, VectorGaussian)

Declaration

public static VectorGaussian BAverageConditional(VectorGaussian product, DistributionArray2D<Gaussian, double> A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussian	product
DistributionArray2D<Gaussian, Double>	A
VectorGaussian	result

Returns

Type	Description
VectorGaussian

BAverageConditional(VectorGaussian, Matrix, VectorGaussian)

EP message to b.

Declaration

public static VectorGaussian BAverageConditional(VectorGaussian product, Matrix A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
Matrix	A	Constant value for `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of b as the random arguments are varied. The formula is proj[p(b) sum_(product) p(product) factor(product,a,b)]/p(b).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

BAverageConditional(VectorGaussian, Double[,], VectorGaussian)

EP message to b.

Declaration

public static VectorGaussian BAverageConditional(VectorGaussian product, double[, ] A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
Double[,]	A	Incoming message from `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of b as the random arguments are varied. The formula is proj[p(b) sum_(product,a) p(product,a) factor(product,a,b)]/p(b).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

BAverageConditional(VectorGaussianMoments, DistributionArray2D<Gaussian, Double>, VectorGaussian)

Declaration

public static VectorGaussian BAverageConditional(VectorGaussianMoments product, DistributionArray2D<Gaussian, double> A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussianMoments	product
DistributionArray2D<Gaussian, Double>	A
VectorGaussian	result

Returns

Type	Description
VectorGaussian

BAverageConditional(VectorGaussianMoments, Matrix, VectorGaussian)

EP message to b.

Declaration

public static VectorGaussian BAverageConditional(VectorGaussianMoments product, Matrix A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussianMoments	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
Matrix	A	Constant value for `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of b as the random arguments are varied. The formula is proj[p(b) sum_(product) p(product) factor(product,a,b)]/p(b).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

BAverageConditional(VectorGaussianMoments, Double[,], VectorGaussian)

EP message to b.

Declaration

public static VectorGaussian BAverageConditional(VectorGaussianMoments product, double[, ] A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussianMoments	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
Double[,]	A	Incoming message from `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of b as the random arguments are varied. The formula is proj[p(b) sum_(product,a) p(product,a) factor(product,a,b)]/p(b).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

BAverageConditional(Vector, Matrix, VectorGaussian)

EP message to b.

Declaration

[NotSupported("A matrix-vector product with fixed output is not yet implemented.")]
public static VectorGaussian BAverageConditional(Vector product, Matrix A, VectorGaussian result)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is the factor viewed as a function of b conditioned on the given values.

BAverageLogarithm(VectorGaussian, Matrix, VectorGaussian)

VMP message to b.

Declaration

public static VectorGaussian BAverageLogarithm(VectorGaussian product, Matrix A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
Matrix	A	Constant value for `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is the factor viewed as a function of b with product integrated out. The formula is sum_product p(product) factor(product,a,b).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

BAverageLogarithm(VectorGaussianMoments, Matrix, VectorGaussian)

VMP message to b.

Declaration

public static VectorGaussian BAverageLogarithm(VectorGaussianMoments product, Matrix A, VectorGaussian result)

Parameters

Type	Name	Description
VectorGaussianMoments	product	Incoming message from `product`. Must be a proper distribution. If any element is uniform, the result will be uniform.
Matrix	A	Constant value for `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is the factor viewed as a function of b with product integrated out. The formula is sum_product p(product) factor(product,a,b).

Exceptions

Type	Condition
ImproperMessageException	`product` is not a proper distribution.

BAverageLogarithm(Vector, Matrix, VectorGaussian)

VMP message to b.

Declaration

[NotSupported("A matrix-vector product with fixed output is not yet implemented.")]
public static VectorGaussian BAverageLogarithm(Vector product, Matrix A, VectorGaussian result)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is the factor viewed as a function of b conditioned on the given values.

BMean(VectorGaussian, PositiveDefiniteMatrix, Vector)

Update the buffer BMean.

Declaration

public static Vector BMean(VectorGaussian B, PositiveDefiniteMatrix BVariance, Vector result)

Parameters

Type	Name	Description
VectorGaussian	B	Incoming message from `b`. Must be a proper distribution. If any element is uniform, the result will be uniform.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
Vector	result	Modified to contain the outgoing message.

Returns

Type	Description
Vector	`result`

Remarks

Exceptions

Type	Condition
ImproperMessageException	`B` is not a proper distribution.

BMeanInit(VectorGaussian)

Initialize the buffer BMean.

Declaration

public static Vector BMeanInit(VectorGaussian B)

Parameters

Type	Name	Description
VectorGaussian	B	Incoming message from `b`.

Returns

Type	Description
Vector	Initial value of buffer `BMean`.

Remarks

BVariance(VectorGaussian, PositiveDefiniteMatrix)

Update the buffer BVariance.

Declaration

public static PositiveDefiniteMatrix BVariance(VectorGaussian B, PositiveDefiniteMatrix result)

Parameters

Type	Name	Description
VectorGaussian	B	Incoming message from `b`. Must be a proper distribution. If any element is uniform, the result will be uniform.
PositiveDefiniteMatrix	result	Modified to contain the outgoing message.

Returns

Type	Description
PositiveDefiniteMatrix	`result`

Remarks

Exceptions

Type	Condition
ImproperMessageException	`B` is not a proper distribution.

BVarianceInit(VectorGaussian)

Initialize the buffer BVariance.

Declaration

public static PositiveDefiniteMatrix BVarianceInit(VectorGaussian B)

Parameters

Type	Name	Description
VectorGaussian	B	Incoming message from `b`.

Returns

Type	Description
PositiveDefiniteMatrix	Initial value of buffer `BVariance`.

Remarks

LogAverageFactor(VectorGaussian, VectorGaussian)

Evidence message for EP.

Declaration

public static double LogAverageFactor(VectorGaussian product, VectorGaussian to_product)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`.
VectorGaussian	to_product	Outgoing message to `product`.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(product) p(product) factor(product,a,b)).

LogAverageFactor(Vector, Matrix, VectorGaussian, Vector, PositiveDefiniteMatrix)

Evidence message for EP.

Declaration

public static double LogAverageFactor(Vector product, Matrix A, VectorGaussian B, Vector BMean, PositiveDefiniteMatrix BVariance)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	B	Incoming message from `b`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(sum_(b) p(b) factor(product,a,b)).

LogAverageFactor(Vector, Matrix, Vector)

Evidence message for EP.

Declaration

public static double LogAverageFactor(Vector product, Matrix A, Vector B)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
Vector	B	Constant value for `b`.

Returns

Type	Description
Double	Logarithm of the factor's average value across the given argument distributions.

Remarks

The formula for the result is log(factor(product,a,b)).

LogEvidenceRatio(VectorGaussian, Matrix, VectorGaussian)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(VectorGaussian product, Matrix A, VectorGaussian B)

Parameters

Type	Name	Description
VectorGaussian	product	Incoming message from `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	B	Incoming message from `b`.

Returns

Type	Description
Double	Logarithm of the factor's contribution the EP model evidence.

Remarks

The formula for the result is log(sum_(product,b) p(product,b) factor(product,a,b) / sum_product p(product) messageTo(product)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

LogEvidenceRatio(Vector, Matrix, VectorGaussian, Vector, PositiveDefiniteMatrix)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(Vector product, Matrix A, VectorGaussian B, Vector BMean, PositiveDefiniteMatrix BVariance)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
VectorGaussian	B	Incoming message from `b`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.

Returns

Type	Description
Double	Logarithm of the factor's contribution the EP model evidence.

Remarks

The formula for the result is log(sum_(b) p(b) factor(product,a,b)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

LogEvidenceRatio(Vector, Matrix, Vector)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(Vector product, Matrix A, Vector B)

Parameters

Type	Name	Description
Vector	product	Constant value for `product`.
Matrix	A	Constant value for `a`.
Vector	B	Constant value for `b`.

Returns

Type	Description
Double	Logarithm of the factor's contribution the EP model evidence.

Remarks

The formula for the result is log(factor(product,a,b)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

ProductAverageConditional(DistributionArray2D<Gaussian, Double>, Vector, PositiveDefiniteMatrix, VectorGaussian)

EP message to product.

Declaration

public static VectorGaussian ProductAverageConditional(DistributionArray2D<Gaussian, double> A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussian result)

Parameters

Type	Name	Description
DistributionArray2D<Gaussian, Double>	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of product as the random arguments are varied. The formula is proj[p(product) sum_(a) p(a) factor(product,a,b)]/p(product).

ProductAverageConditional(DistributionArray2D<Gaussian, Double>, Vector, PositiveDefiniteMatrix, VectorGaussianMoments)

EP message to product.

Declaration

public static VectorGaussianMoments ProductAverageConditional(DistributionArray2D<Gaussian, double> A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussianMoments result)

Parameters

Type	Name	Description
DistributionArray2D<Gaussian, Double>	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussianMoments	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussianMoments	`result`

Remarks

The outgoing message is a distribution matching the moments of product as the random arguments are varied. The formula is proj[p(product) sum_(a) p(a) factor(product,a,b)]/p(product).

ProductAverageConditional(Matrix, Vector, PositiveDefiniteMatrix, VectorGaussian)

EP message to product.

Declaration

public static VectorGaussian ProductAverageConditional(Matrix A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussian result)

Parameters

Type	Name	Description
Matrix	A	Constant value for `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is the factor viewed as a function of product conditioned on the given values.

ProductAverageConditional(Matrix, Vector, PositiveDefiniteMatrix, VectorGaussianMoments)

EP message to product.

Declaration

public static VectorGaussianMoments ProductAverageConditional(Matrix A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussianMoments result)

Parameters

Type	Name	Description
Matrix	A	Constant value for `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussianMoments	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussianMoments	`result`

Remarks

The outgoing message is the factor viewed as a function of product conditioned on the given values.

ProductAverageConditional(Double[,], Vector, PositiveDefiniteMatrix, VectorGaussian)

EP message to product.

Declaration

public static VectorGaussian ProductAverageConditional(double[, ] A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussian result)

Parameters

Type	Name	Description
Double[,]	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of product as the random arguments are varied. The formula is proj[p(product) sum_(a) p(a) factor(product,a,b)]/p(product).

ProductAverageConditional(Double[,], Vector, PositiveDefiniteMatrix, VectorGaussianMoments)

EP message to product.

Declaration

public static VectorGaussianMoments ProductAverageConditional(double[, ] A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussianMoments result)

Parameters

Type	Name	Description
Double[,]	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussianMoments	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussianMoments	`result`

Remarks

The outgoing message is a distribution matching the moments of product as the random arguments are varied. The formula is proj[p(product) sum_(a) p(a) factor(product,a,b)]/p(product).

ProductAverageConditionalInit(Matrix)

Declaration

public static VectorGaussian ProductAverageConditionalInit(Matrix A)

Parameters

Type	Name	Description
Matrix	A	Constant value for `a`.

Returns

Type	Description
VectorGaussian

Remarks

ProductAverageLogarithm(Matrix, Vector, PositiveDefiniteMatrix, VectorGaussian)

VMP message to product.

Declaration

public static VectorGaussian ProductAverageLogarithm(Matrix A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussian result)

Parameters

Type	Name	Description
Matrix	A	Constant value for `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is the factor viewed as a function of product conditioned on the given values.

ProductAverageLogarithm(Matrix, Vector, PositiveDefiniteMatrix, VectorGaussianMoments)

VMP message to product.

Declaration

public static VectorGaussianMoments ProductAverageLogarithm(Matrix A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussianMoments result)

Parameters

Type	Name	Description
Matrix	A	Constant value for `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussianMoments	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussianMoments	`result`

Remarks

The outgoing message is the factor viewed as a function of product conditioned on the given values.

ProductAverageLogarithm(Double[,], Vector, PositiveDefiniteMatrix, VectorGaussian)

VMP message to product.

Declaration

public static VectorGaussian ProductAverageLogarithm(double[, ] A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussian result)

Parameters

Type	Name	Description
Double[,]	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussian	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussian	`result`

Remarks

The outgoing message is a distribution matching the moments of product as the random arguments are varied. The formula is proj[sum_(a) p(a) factor(product,a,b)].

ProductAverageLogarithm(Double[,], Vector, PositiveDefiniteMatrix, VectorGaussianMoments)

VMP message to product.

Declaration

public static VectorGaussianMoments ProductAverageLogarithm(double[, ] A, Vector BMean, PositiveDefiniteMatrix BVariance, VectorGaussianMoments result)

Parameters

Type	Name	Description
Double[,]	A	Incoming message from `a`.
Vector	BMean	Buffer `BMean`.
PositiveDefiniteMatrix	BVariance	Buffer `BVariance`.
VectorGaussianMoments	result	Modified to contain the outgoing message.

Returns

Type	Description
VectorGaussianMoments	`result`

Remarks

The outgoing message is a distribution matching the moments of product as the random arguments are varied. The formula is proj[sum_(a) p(a) factor(product,a,b)].

ProductAverageLogarithmInit(Matrix)

Declaration

public static VectorGaussian ProductAverageLogarithmInit(Matrix A)

Parameters

Type	Name	Description
Matrix	A	Constant value for `a`.

Returns

Type	Description
VectorGaussian