Class GetDeepJaggedItemsOp<T>

Provides outgoing messages for GetDeepJaggedItems<T>(IReadOnlyList<T>, IReadOnlyList<IReadOnlyList<IReadOnlyList<Int32>>>), given random arguments to the function.

Inheritance

Object

GetDeepJaggedItemsOp<T>

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(Collection), "GetDeepJaggedItems<>", new Type[]{}, Default = true)]
[Quality(QualityBand.Mature)]
[Buffers(new string[]{"marginal"})]
public static class GetDeepJaggedItemsOp<T>

Type Parameters

Name	Description
T	The type of an item.

Methods

ArrayAverageConditional<DistributionType, ArrayType>(IList<IList<IList<T>>>, IList<IList<IList<Int32>>>, ArrayType)

EP message to array.

Declaration

public static ArrayType ArrayAverageConditional<DistributionType, ArrayType>([SkipIfAllUniform] IList<IList<IList<T>>> items, IList<IList<IList<int>>> indices, ArrayType result)
    where DistributionType : HasPoint<T> where ArrayType : IList<DistributionType>, SettableToUniform

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items. Must be a proper distribution. If all elements are uniform, the result will be uniform.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.
ArrayType	result	Modified to contain the outgoing message.

Returns

Type	Description
ArrayType	result

Type Parameters

Name	Description
DistributionType	The type of a distribution over array elements.
ArrayType	The type of the resulting array.

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	items is not a proper distribution.

ArrayAverageConditional<ItemType, ItemType2, DistributionType, ArrayType>(IList<ItemType>, IList<IList<IList<Int32>>>, ArrayType)

Declaration

public static ArrayType ArrayAverageConditional<ItemType, ItemType2, DistributionType, ArrayType>([SkipIfAllUniform] IList<ItemType> items, IList<IList<IList<int>>> indices, ArrayType result)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableToProduct<DistributionType> where ArrayType : IList<DistributionType>, SettableToUniform

Parameters

Type	Name	Description
IList<ItemType>	items
IList<IList<IList<Int32>>>	indices
ArrayType	result

Returns

Type	Description
ArrayType

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.
ArrayType	The type of the resulting array.

ArrayAverageLogarithm<DistributionType, ArrayType>(IList<IList<IList<T>>>, IList<IList<IList<Int32>>>, ArrayType)

VMP message to array.

Declaration

public static ArrayType ArrayAverageLogarithm<DistributionType, ArrayType>(IList<IList<IList<T>>> items, IList<IList<IList<int>>> indices, ArrayType result)
    where DistributionType : HasPoint<T> where ArrayType : IList<DistributionType>, SettableToUniform

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.
ArrayType	result	Modified to contain the outgoing message.

Returns

Type	Description
ArrayType	result

Type Parameters

Name	Description
DistributionType	The type of a distribution over array elements.
ArrayType	The type of the resulting array.

Remarks

The outgoing message is the exponential of the average log-factor value, where the average is over all arguments except array. Because the factor is deterministic, items is integrated out before taking the logarithm. The formula is exp(sum_(indices) p(indices) log(sum_items p(items) factor(items,array,indices))).

ArrayAverageLogarithm<ItemType, ItemType2, DistributionType, ArrayType>(IList<ItemType>, IList<IList<IList<Int32>>>, ArrayType)

VMP message to array.

Declaration

public static ArrayType ArrayAverageLogarithm<ItemType, ItemType2, DistributionType, ArrayType>([SkipIfAllUniform] IList<ItemType> items, IList<IList<IList<int>>> indices, ArrayType result)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableToUniform, SettableToProduct<DistributionType> where ArrayType : IList<DistributionType>, SettableToUniform

Parameters

Type	Name	Description
IList<ItemType>	items	Incoming message from items. Must be a proper distribution. If all elements are uniform, the result will be uniform.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.
ArrayType	result	Modified to contain the outgoing message.

Returns

Type	Description
ArrayType	result

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.
ArrayType	The type of the resulting array.

Remarks

The outgoing message is the exponential of the average log-factor value, where the average is over all arguments except array. Because the factor is deterministic, items is integrated out before taking the logarithm. The formula is exp(sum_(indices) p(indices) log(sum_items p(items) factor(items,array,indices))).

Exceptions

Type	Condition
ImproperMessageException	items is not a proper distribution.

AverageLogFactor(IList<IList<IList<T>>>, IList<T>, IList<IList<IList<Int32>>>)

Evidence message for VMP.

Declaration

public static double AverageLogFactor(IList<IList<IList<T>>> items, IList<T> array, IList<IList<IList<int>>> indices)

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<T>	array	Incoming message from array.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.

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<DistributionType>(IList<IList<IList<T>>>, IList<DistributionType>)

Evidence message for VMP.

Declaration

public static double AverageLogFactor<DistributionType>(IList<IList<IList<T>>> items, IList<DistributionType> array)
    where DistributionType : HasPoint<T>, CanGetLogProb<T>

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<DistributionType>	array	Incoming message from array.

Returns

Type	Description
Double	Zero.

Type Parameters

Name	Description
DistributionType	The type of a distribution over array elements.

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<ItemType, ItemType2, DistributionType>(IList<ItemType>, IList<DistributionType>)

Evidence message for VMP.

Declaration

public static double AverageLogFactor<ItemType, ItemType2, DistributionType>(IList<ItemType> items, IList<DistributionType> array)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableToProduct<DistributionType>, CanGetLogAverageOf<DistributionType>

Parameters

Type	Name	Description
IList<ItemType>	items	Incoming message from items.
IList<DistributionType>	array	Incoming message from array.

Returns

Type	Description
Double	Zero.

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

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<ItemType, ItemType2, DistributionType>(IList<ItemType>, IList<T>)

Declaration

public static double AverageLogFactor<ItemType, ItemType2, DistributionType>(IList<ItemType> items, IList<T> array)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : CanGetLogProb<T>

Parameters

Type	Name	Description
IList<ItemType>	items
IList<T>	array

Returns

Type	Description
Double

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

ItemsAverageConditional<ArrayType, ItemType, ItemType2, DistributionType>(ItemType, ArrayType, ArrayType, IList<IList<IList<Int32>>>, Int32, ItemType)

EP message to items.

Declaration

public static ItemType ItemsAverageConditional<ArrayType, ItemType, ItemType2, DistributionType>(ItemType items, ArrayType array, [SkipIfAllUniform] ArrayType marginal, IList<IList<IList<int>>> indices, int resultIndex, ItemType result)
    where ArrayType : IList<DistributionType> where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableToRatio<DistributionType>

Parameters

Type	Name	Description
ItemType	items	Incoming message from items.
ArrayType	array	Incoming message from array.
ArrayType	marginal	Buffer marginal.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.
Int32	resultIndex	Index of the items for which a message is desired.
ItemType	result	Modified to contain the outgoing message.

Returns

Type	Description
ItemType	result

Type Parameters

Name	Description
ArrayType	The type of an array for the marginal.
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

Remarks

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

ItemsAverageLogarithm<ItemType, ItemType2, DistributionType>(IList<DistributionType>, IList<IList<IList<Int32>>>, Int32, ItemType)

VMP message to items.

Declaration

public static ItemType ItemsAverageLogarithm<ItemType, ItemType2, DistributionType>([SkipIfAllUniform] IList<DistributionType> array, IList<IList<IList<int>>> indices, int resultIndex, ItemType result)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableTo<DistributionType>

Parameters

Type	Name	Description
IList<DistributionType>	array	Incoming message from array. Must be a proper distribution. If all elements are uniform, the result will be uniform.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.
Int32	resultIndex	Index of the items for which a message is desired.
ItemType	result	Modified to contain the outgoing message.

Returns

Type	Description
ItemType	result

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

Remarks

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

Exceptions

Type	Condition
ImproperMessageException	array is not a proper distribution.

LogAverageFactor(IList<IList<IList<T>>>, IList<T>, IList<IList<IList<Int32>>>)

Evidence message for EP.

Declaration

public static double LogAverageFactor(IList<IList<IList<T>>> items, IList<T> array, IList<IList<IList<int>>> indices)

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<T>	array	Incoming message from array.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.

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_(items,array,indices) p(items,array,indices) factor(items,array,indices)).

LogAverageFactor<DistributionType>(IList<IList<IList<T>>>, IList<DistributionType>, IList<IList<IList<Int32>>>)

Evidence message for EP.

Declaration

public static double LogAverageFactor<DistributionType>(IList<IList<IList<T>>> items, IList<DistributionType> array, IList<IList<IList<int>>> indices)
    where DistributionType : HasPoint<T>, CanGetLogProb<T>

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<DistributionType>	array	Incoming message from array.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.

Returns

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

Type Parameters

Name	Description
DistributionType	The type of a distribution over array elements.

Remarks

The formula for the result is log(sum_(items,array,indices) p(items,array,indices) factor(items,array,indices)).

LogAverageFactor<ItemType, ItemType2, DistributionType>(IList<ItemType>, IList<DistributionType>, IList<IList<IList<Int32>>>)

Declaration

public static double LogAverageFactor<ItemType, ItemType2, DistributionType>(IList<ItemType> items, IList<DistributionType> array, IList<IList<IList<int>>> indices)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : IDistribution<T>, SettableToProduct<DistributionType>, CanGetLogAverageOf<DistributionType>

Parameters

Type	Name	Description
IList<ItemType>	items
IList<DistributionType>	array
IList<IList<IList<Int32>>>	indices

Returns

Type	Description
Double

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

LogAverageFactor<ItemType, ItemType2, DistributionType>(IList<ItemType>, IList<T>, IList<IList<IList<Int32>>>)

Evidence message for EP.

Declaration

public static double LogAverageFactor<ItemType, ItemType2, DistributionType>(IList<ItemType> items, IList<T> array, IList<IList<IList<int>>> indices)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : CanGetLogProb<T>

Parameters

Type	Name	Description
IList<ItemType>	items	Incoming message from items.
IList<T>	array	Incoming message from array.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.

Returns

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

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

Remarks

The formula for the result is log(sum_(items,array,indices) p(items,array,indices) factor(items,array,indices)).

LogEvidenceRatio(IList<IList<IList<T>>>, IList<T>, IList<IList<IList<Int32>>>)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio(IList<IList<IList<T>>> items, IList<T> array, IList<IList<IList<int>>> indices)

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<T>	array	Incoming message from array.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.

Returns

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

Remarks

The formula for the result is log(sum_(items,array,indices) p(items,array,indices) factor(items,array,indices) / sum_items p(items) messageTo(items)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

LogEvidenceRatio<DistributionType>(IList<IList<IList<T>>>, IList<DistributionType>, IList<IList<IList<Int32>>>)

Evidence message for EP.

Declaration

public static double LogEvidenceRatio<DistributionType>(IList<IList<IList<T>>> items, IList<DistributionType> array, IList<IList<IList<int>>> indices)
    where DistributionType : HasPoint<T>, CanGetLogProb<T>

Parameters

Type	Name	Description
IList<IList<IList<T>>>	items	Incoming message from items.
IList<DistributionType>	array	Incoming message from array.
IList<IList<IList<Int32>>>	indices	Incoming message from indices.

Returns

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

Type Parameters

Name	Description
DistributionType	The type of a distribution over array elements.

Remarks

The formula for the result is log(sum_(items,array,indices) p(items,array,indices) factor(items,array,indices) / sum_items p(items) messageTo(items)). Adding up these values across all factors and variables gives the log-evidence estimate for EP.

LogEvidenceRatio<ItemType, ItemType2, DistributionType>(IList<ItemType>, IList<DistributionType>, IList<IList<IList<Int32>>>, IList<ItemType>)

Declaration

public static double LogEvidenceRatio<ItemType, ItemType2, DistributionType>(IList<ItemType> items, IList<DistributionType> array, IList<IList<IList<int>>> indices, IList<ItemType> to_items)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableToUniform, SettableToProduct<DistributionType>, CanGetLogAverageOf<DistributionType>, ICloneable

Parameters

Type	Name	Description
IList<ItemType>	items
IList<DistributionType>	array
IList<IList<IList<Int32>>>	indices
IList<ItemType>	to_items

Returns

Type	Description
Double

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

LogEvidenceRatio<ItemType, ItemType2, DistributionType>(IList<ItemType>, IList<T>, IList<IList<IList<Int32>>>)

Declaration

public static double LogEvidenceRatio<ItemType, ItemType2, DistributionType>(IList<ItemType> items, IList<T> array, IList<IList<IList<int>>> indices)
    where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : CanGetLogProb<T>

Parameters

Type	Name	Description
IList<ItemType>	items
IList<T>	array
IList<IList<IList<Int32>>>	indices

Returns

Type	Description
Double

Type Parameters

Name	Description
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

Marginal<ArrayType, DistributionType>(ArrayType, ArrayType, ArrayType)

Declaration

[SkipIfAllUniform(new string[]{"array", "to_array"})]
public static ArrayType Marginal<ArrayType, DistributionType>(ArrayType array, ArrayType to_array, ArrayType result)
    where ArrayType : IList<DistributionType>, SettableTo<ArrayType> where DistributionType : SettableToProduct<DistributionType>

Parameters

Type	Name	Description
ArrayType	array
ArrayType	to_array
ArrayType	result

Returns

Type	Description
ArrayType

Type Parameters

Name	Description
ArrayType	The type of an array for the marginal.
DistributionType	The type of a distribution over array elements.

MarginalIncrement<ArrayType, ItemType, ItemType2, DistributionType>(ArrayType, ItemType, ItemType, IList<IList<IList<Int32>>>, Int32)

Declaration

public static ArrayType MarginalIncrement<ArrayType, ItemType, ItemType2, DistributionType>(ArrayType result, ItemType to_item, ItemType item, IList<IList<IList<int>>> indices, int resultIndex)
    where ArrayType : IList<DistributionType>, SettableTo<ArrayType> where ItemType : IList<ItemType2> where ItemType2 : IList<DistributionType> where DistributionType : SettableToProduct<DistributionType>

Parameters

Type	Name	Description
ArrayType	result	Modified to contain the outgoing message.
ItemType	to_item
ItemType	item
IList<IList<IList<Int32>>>	indices	Incoming message from indices.
Int32	resultIndex	Index of the marginal for which a message is desired.

Returns

Type	Description
ArrayType	result

Type Parameters

Name	Description
ArrayType	The type of an array for the marginal.
ItemType	The type of a sub-array.
ItemType2	The type of a sub-sub-array.
DistributionType	The type of a distribution over array elements.

Remarks

MarginalInit<ArrayType>(ArrayType)

Initialize the buffer marginal.

Declaration

public static ArrayType MarginalInit<ArrayType>(ArrayType array)
    where ArrayType : ICloneable

Parameters

Type	Name	Description
ArrayType	array	Incoming message from array. Must be a proper distribution. If any element is uniform, the result will be uniform.

Returns

Type	Description
ArrayType	Initial value of buffer marginal.

Type Parameters

Name	Description
ArrayType	The type of an array for the marginal.

Remarks

Exceptions

Type	Condition
ImproperMessageException	array is not a proper distribution.