Rules implementation
Message update rules
ReactiveMP.rule
— Functionrule(fform, on, vconstraint, mnames, messages, qnames, marginals, meta, __node)
This function is used to compute an outbound message for a given node
Arguments
fform
: Functional form of the node in form of a type of the node, e.g.::Type{ <: NormalMeanVariance }
or::typeof(+)
on
: Outbound interface's tag for which a message has to be computed, e.g.::Val{:out}
or::Val{:μ}
vconstraint
: Variable constraints for an outbound interface, e.g.Marginalisation
orMomentMatching
mnames
: Ordered messages names in form of the Val type, eg.::Val{ (:mean, :precision) }
messages
: Tuple of message of the same length asmnames
used to compute an outbound messageqnames
: Ordered marginal names in form of the Val type, eg.::Val{ (:mean, :precision) }
marginals
: Tuple of marginals of the same length asqnames
used to compute an outbound messagemeta
: Extra meta informationaddons
: Extra addons information__node
: Node reference
For all available rules, see ReactiveMP.print_rules_table()
.
See also: @rule
, marginalrule
, @marginalrule
ReactiveMP.@rule
— Macro@rule NodeType(:Edge, Constraint) (Arguments..., [ meta::MetaType ]) = begin
# rule body
return ...
end
The @rule
macro help to define new methods for the rule
function. It works particularly well in combination with the @node
macro. It has a specific structure, which must specify:
NodeType
: must be a valid Julia type. If some attempt to define a rule for a Julia function (for example+
), usetypeof(+)
Edge
: edge label, usually edge labels are defined with the@node
macroConstrain
: DEPRECATED, please just use theMarginalisation
labelArguments
: defines a list of the input arguments for the rulem_*
prefix indicates that the argument is of typeMessage
from the edge*
q_*
prefix indicates that the argument is of typeMarginal
on the edge*
Meta::MetaType
- optionally, a user can specify aMeta
object of typeMetaType
. This can be useful is some attempts to try different rules with different approximation methods or if the rule itself requires some temporary storage or cache. The default meta isnothing
.
Here are various examples of the @rule
macro usage:
- Belief-Propagation (or Sum-Product) message update rule for the
NormalMeanVariance
node toward the:μ
edge with theMarginalisation
constraint. Input arguments arem_out
andm_v
, which are the messages from the corresponding edgesout
andv
and have the typePointMass
.
@rule NormalMeanVariance(:μ, Marginalisation) (m_out::PointMass, m_v::PointMass) = NormalMeanVariance(mean(m_out), mean(m_v))
- Mean-field message update rule for the
NormalMeanVariance
node towards the:μ
edge with theMarginalisation
constraint. Input arguments areq_out
andq_v
, which are the marginals on the corresponding edgesout
andv
of typeAny
.
@rule NormalMeanVariance(:μ, Marginalisation) (q_out::Any, q_v::Any) = NormalMeanVariance(mean(q_out), mean(q_v))
- Structured Variational message update rule for the
NormalMeanVariance
node towards the:out
edge with theMarginalisation
constraint. Input arguments arem_μ
, which is a message from theμ
edge of typeUnivariateNormalDistributionsFamily
, andq_v
, which is a marginal on thev
edge of typeAny
.
@rule NormalMeanVariance(:out, Marginalisation) (m_μ::UnivariateNormalDistributionsFamily, q_v::Any) = begin
m_μ_mean, m_μ_cov = mean_cov(m_μ)
return NormalMeanVariance(m_μ_mean, m_μ_cov + mean(q_v))
end
See also: rule
, marginalrule
, [@marginalrule
], @call_rule
ReactiveMP.@call_rule
— Macro@call_rule NodeType(:edge, Constraint) (argument1 = value1, argument2 = value2, ..., [ meta = ..., addons = ... ])
The @call_rule
macro helps to call the rule
method with an easier syntax. The structure of the macro is almost the same as in the @rule
macro, but there is no begin ... end
block, but instead each argument must have a specified value with the =
operator.
The @call_rule
accepts optional list of options before the functional form specification, for example:
@call_rule [ return_addons = true ] NodeType(:edge, Constraint) (argument1 = value1, argument2 = value2, ..., [ meta = ..., addons = ... ])
The list of available options is:
return_addons
- forces the@call_rule
to return the tuple of(result, addons)
fallback
- specifies the fallback rule to use in case the rule is not defined for the givenNodeType
and specified arguments
See also: @rule
, rule
, @call_marginalrule
ReactiveMP.call_rule_make_node
— Functioncall_rule_create_node(::Type{ NodeType }, fformtype)
Creates a node object that will be used inside @call_rule
macro.
ReactiveMP.call_rule_macro_parse_fn_args
— Functioncall_rule_macro_parse_fn_args(inputs; specname, prefix, proxy)
Do not use this function directly. This function is private and does not belong to the public API.
This function is used to parse an arguments
tuple for message and marginal calling rules specification.
@call_rule MvNormalMeanPrecision(:out, Marginalisation) (m_μ = NormalMeanPrecision(...), m_τ = PointMass(...)) = begin
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
`arguments` vector
...
end
Accepts a vector of (name, value) elements, specname, name prefix and proxy type. Returns parsed names without prefix and proxied values
See also: @rule
ReactiveMP.call_rule_is_node_required
— Functioncall_rule_is_node_required(fformtype)
Returns either CallRuleNodeRequired()
or CallRuleNodeNotRequired()
depending on if a specific fformtype
requires an access to the corresponding node in order to compute a message update rule. Returns CallRuleNodeNotRequired()
for all known functional forms by default and CallRuleNodeRequired()
for all unknown functional forms.
ReactiveMP.rule_macro_parse_on_tag
— Functionrule_macro_parse_on_tag(expression)
Do not use this function directly. This function is private and does not belong to the public API.
This function is used to parse an on
tag for message rules and marginal rules specification.
@rule MvNormalMeanPrecision(:out, Marginalisation) (...) = begin
^^^^
`on` tag
...
end
or
@rule NormalMixture((:m, k), Marginalisation) (...) = begin
^^^^^^^
`on` tag
...
end
Accepts either a quoted symbol expressions or a (name, index) tuple expression. Returns name expression, index expression and index initialisation expression.
See also: @rule
ReactiveMP.rule_macro_parse_fn_args
— Functionrule_macro_parse_fn_args(inputs; specname, prefix, proxy)
Do not use this function directly. This function is private and does not belong to the public API.
This function is used to parse an arguments
tuple for message rules and marginal rules specification.
@rule MvNormalMeanPrecision(:out, Marginalisation) (m_μ::NormalMeanPrecision, m_τ::PointMass) = begin
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
`arguments` vector
...
end
Accepts a vector of (name, type) elements, specname, name prefix and proxy type. Returns parsed names without prefix, proxied types and initialisation code block.
See also: @rule
ReactiveMP.rule_macro_check_fn_args
— Functionrule_macro_check_fn_args(inputs; allowed_inputs, allowed_prefixes)
This function checks if all inputs
are either in the allowed_inputs
or have prefixes in the allowed_prefixes
.
See also: @rule
Marginal update rules
ReactiveMP.marginalrule
— Functionmarginalrule(fform, on, mnames, messages, qnames, marginals, meta, __node)
This function is used to compute a local joint marginal for a given node
Arguments
fform
: Functional form of the node in form of a type of the node, e.g.::Type{ <: NormalMeanVariance }
or::typeof(+)
on
: Local joint marginal tag , e.g.::Val{ :mean_precision }
or::Val{ :out_mean_precision }
mnames
: Ordered messages names in form of the Val type, eg.::Val{ (:mean, :precision) }
messages
: Tuple of message of the same length asmnames
used to compute an outbound messageqnames
: Ordered marginal names in form of the Val type, eg.::Val{ (:mean, :precision) }
marginals
: Tuple of marginals of the same length asqnames
used to compute an outbound messagemeta
: Extra meta information__node
: Node reference
See also: rule
, @rule
@marginalrule
ReactiveMP.@marginalrule
— Macro@marginalrule NodeType(:Cluster) (Arguments..., [ meta::MetaType ]) = begin
# rule body
return ...
end
The @marginalrule
macro help to define new methods for the marginalrule
function. It works particularly well in combination with the @node
macro. It has a specific structure, which must specify:
NodeType
: must be a valid Julia type. If some attempt to define a rule for a Julia function (for example+
), usetypeof(+)
Cluster
: edge cluster that contains joined edge labels with the_
symbol. Usually edge labels are defined with the@node
macroArguments
: defines a list of the input arguments for the rulem_*
prefix indicates that the argument is of typeMessage
from the edge*
q_*
prefix indicates that the argument is of typeMarginal
on the edge*
Meta::MetaType
- optionally, a user can specify aMeta
object of typeMetaType
. This can be useful is some attempts to try different rules with different approximation methods or if the rule itself requires some temporary storage or cache. The default meta isnothing
.
The @marginalrule
can return a NamedTuple
in the return
statement. This would indicate some variables in the joint marginal for the Cluster
are independent and the joint itself is factorised. For example if some attempts to compute a marginal for the q(x, y)
it is possible to return (x = ..., y = ...)
as the result of the computation to indicate that q(x, y) = q(x)q(y)
.
Here are various examples of the @marginalrule
macro usage:
- Marginal computation rule around the
NormalMeanPrecision
node for theq(out, μ)
. The rule accepts argumentsm_out
andm_μ
, which are the messages
from the out
and μ
edges respectively, and q_τ
which is the marginal on the edge τ
.
@marginalrule NormalMeanPrecision(:out_μ) (m_out::UnivariateNormalDistributionsFamily, m_μ::UnivariateNormalDistributionsFamily, q_τ::Any) = begin
xi_out, W_out = weightedmean_precision(m_out)
xi_μ, W_μ = weightedmean_precision(m_μ)
W_bar = mean(q_τ)
W = [W_out+W_bar -W_bar; -W_bar W_μ+W_bar]
xi = [xi_out; xi_μ]
return MvNormalWeightedMeanPrecision(xi, W)
end
- Marginal computation rule around the
NormalMeanPrecision
node for theq(out, μ)
. The rule accepts argumentsm_out
andm_μ
, which are the messages from the
out
and μ
edges respectively, and q_τ
which is the marginal on the edge τ
. In this example the result of the computation is a NamedTuple
@marginalrule NormalMeanPrecision(:out_μ) (m_out::PointMass, m_μ::UnivariateNormalDistributionsFamily, q_τ::Any) = begin
return (out = m_out, μ = prod(ClosedProd(), NormalMeanPrecision(mean(m_out), mean(q_τ)), m_μ))
end
ReactiveMP.@call_marginalrule
— Macro@call_marginalrule NodeType(:edge) (argument1 = value1, argument2 = value2, ..., [ meta = ... ])
The @call_marginalrule
macro helps to call the marginalrule
method with an easier syntax. The structure of the macro is almost the same as in the @marginalrule
macro, but there is no begin ... end
block, but instead each argument must have a specified value with the =
operator.
See also: @marginalrule
, marginalrule
, @call_rule
Testing utilities for the update rules
ReactiveMP.@test_rules
— Macro@test_rules [options] rule [ test_entries... ]
The @test_rules
macro generates test cases for message update rules for probabilistic programming models that follow the "message passing" paradigm. It takes a rule specification as input and generates a set of tests based on that specification. This macro is provided by ReactiveMP
.
Note: The Test
module must be imported explicitly. The @test_rules
macro tries to use the @test
macro, which must be defined globally.
Arguments
The macro takes three arguments:
options
: An optional argument that specifies the options for the test generation process. See below for details.rule
: A rule specification in the same format as the@rule
macro, e.g.Beta(:out, Marginalisation)
orNormalMeanVariance(:μ, Marginalisation)
.test_entries
: An array of named tuples(input = ..., output = ...)
. Theinput
entry has the same format as the input for the@rule
macro. Theoutput
entry specifies the expected output.
Options
The following options are available:
check_type_promotion
: By default, this option is set tofalse
. If set totrue
, the macro generates an extensive list of extra tests that aim to check the correct type promotion within the tests. For example, if all inputs are of typeFloat32
, then the expected output should also be of typeFloat32
. See theparamfloattype
andconvert_paramfloattype
functions for details.atol
: Sets the desired accuracy for the tests. The tests use thecustom_rule_isapprox
function fromReactiveMP
to check if outputs are approximately the same. This argument can be either a single number or an array ofkey => value
pairs.extra_float_types
: A set of extra float types to be used in thecheck_type_promotion
tests. This argument has no effect ifcheck_type_promotion
is set tofalse
.
The default values for the atol
option are:
Float32
:1e-4
Float64
:1e-6
BigFloat
:1e-8
Examples
@test_rules [check_type_promotion = true] Beta(:out, Marginalisation) [
(input = (m_a = PointMass(1.0), m_b = PointMass(2.0)), output = Beta(1.0, 2.0)),
(input = (m_a = PointMass(2.0), m_b = PointMass(2.0)), output = Beta(2.0, 2.0)),
(input = (m_a = PointMass(3.0), m_b = PointMass(3.0)), output = Beta(3.0, 3.0))
]
@test_rules [check_type_promotion = true] Beta(:out, Marginalisation) [
(input = (q_a = PointMass(1.0), q_b = PointMass(2.0)), output = Beta(1.0, 2.0)),
(input = (q_a = PointMass(2.0), q_b = PointMass(2.0)), output = Beta(2.0, 2.0)),
(input = (q_a = PointMass(3.0), q_b = PointMass(3.0)), output = Beta(3.0, 3.0))
]
See also: ReactiveMP.@test_marginalrules
ReactiveMP.@test_marginalrules
— Macro@test_marginalrules [options] rule [ test_entries... ]
Effectively the same as @test_rules
, but for marginal computation rules. See the documentation for @test_rules
for more info.
See also: ReactiveMP.@test_rules
Rule fallbacks
ReactiveMP.NodeFunctionRuleFallback
— TypeNodeFunctionRuleFallback(extractfn = mean)
A fallback rule for Stochastic
nodes that uses a specified function (default: mean) to transform messages and marginals into a value. It calls the nodefunction
method to create the message.
When a node is defined with the @node
macro:
1. The nodefunction
typically calls logpdf
associated with the node's distribution. 2. The first edge in the @node
specification is used to evaluate logpdf
at. 3. Other edges are used to instantiate the associated distribution object.
julia> using ReactiveMP, BayesBase, Distributions
julia> struct MyBeta{A, B} <: ContinuousUnivariateDistribution
a::A
b::B
end
julia> BayesBase.logpdf(d::MyBeta, x) = logpdf(Beta(d.a, d.b), x)
julia> BayesBase.insupport(d::MyBeta, x::Real) = true
julia> @node MyBeta Stochastic [out, a, b]
julia> message = @call_rule [fallback = NodeFunctionRuleFallback()] MyBeta(:out, Marginalisation) (m_a = Beta(2, 3), m_b = Beta(3, 2));
julia> logpdf(message, 0.5)
-0.5017644952110732
julia> message = @call_rule [fallback = NodeFunctionRuleFallback(mode)] MyBeta(:out, Marginalisation) (m_a = Beta(2, 3), m_b = Beta(3, 2)); # evaluate at `mode`
julia> logpdf(message, 0.5)
-0.5954237415153454
ReactiveMP.nodefunction
— Functionnodefunction(::Type{T}) where {T}
Returns a function that represents a node of type T
. The function typically takes arguments that represent the node's input and output variables in the same order as defined in the @node
macro.
Table of available update rules
The list below has been automatically generated with the ReactiveMP.print_rules_table()
function.
Node | Output | Inputs | Meta |
---|---|---|---|
dot | in2 | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in1) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
dot | in2 | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in1) :: BayesBase.PointMass | |||
dot | in1 | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in2) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
dot | in1 | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in2) :: BayesBase.PointMass | |||
dot | out | μ(in1) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in2) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
dot | out | μ(in1) :: BayesBase.PointMass | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in2) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
dot | out | μ(in1) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: BayesBase.PointMass | Nothing |
μ(in2) :: ExponentialFamily.MvNormalWeightedMeanPrecision | |||
+ | out | μ(in1) :: ExponentialFamily.MvNormalWeightedMeanPrecision | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: ExponentialFamily.MvNormalWeightedMeanPrecision{T1, M, P} where {M<:AbstractVector{T1}, P<:AbstractMatrix{T1}} | Nothing |
μ(in2) :: ExponentialFamily.MvNormalWeightedMeanPrecision{T2, M, P} where {M<:AbstractVector{T2}, P<:AbstractMatrix{T2}} | |||
+ | out | μ(in1) :: BayesBase.PointMass | Nothing |
μ(in2) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
+ | out | μ(in1) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: BayesBase.PointMass | Nothing |
μ(in2) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
+ | out | μ(in1) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: BayesBase.PointMass | Nothing |
μ(in2) :: ExponentialFamily.MvNormalMeanPrecision | |||
+ | out | μ(in1) :: ExponentialFamily.MvNormalMeanPrecision | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: BayesBase.PointMass | Nothing |
μ(in2) :: ExponentialFamily.NormalMeanPrecision | |||
+ | out | μ(in1) :: ExponentialFamily.NormalMeanPrecision | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: BayesBase.PointMass | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | out | μ(in1) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
+ | out | μ(in1) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
+ | out | μ(in1) :: Distributions.Distribution | Nothing |
μ(in2) :: Distributions.Distribution | |||
+ | in2 | μ(out) :: Any | |
μ(in1) :: Any | |||
- | in2 | μ(out) :: Any | |
μ(in1) :: Any | |||
- | in1 | μ(out) :: Any | |
μ(in2) :: Any | |||
- | out | μ(in1) :: Any | |
μ(in2) :: Any | |||
+ | in1 | μ(out) :: BayesBase.PointMass | Nothing |
μ(in2) :: ExponentialFamily.MvNormalWeightedMeanPrecision | |||
+ | in1 | μ(out) :: ExponentialFamily.MvNormalWeightedMeanPrecision{T1, M, P} where {M<:AbstractVector{T1}, P<:AbstractMatrix{T1}} | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | in1 | μ(out) :: ExponentialFamily.MvNormalWeightedMeanPrecision{T1, M, P} where {M<:AbstractVector{T1}, P<:AbstractMatrix{T1}} | Nothing |
μ(in2) :: ExponentialFamily.MvNormalWeightedMeanPrecision{T2, M, P} where {M<:AbstractVector{T2}, P<:AbstractMatrix{T2}} | |||
+ | in1 | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | in1 | μ(out) :: BayesBase.PointMass | Nothing |
μ(in2) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
+ | in1 | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | in1 | μ(out) :: BayesBase.PointMass | Nothing |
μ(in2) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
+ | in1 | μ(out) :: ExponentialFamily.MvNormalMeanPrecision | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | in1 | μ(out) :: BayesBase.PointMass | Nothing |
μ(in2) :: ExponentialFamily.MvNormalMeanPrecision | |||
+ | in1 | μ(out) :: ExponentialFamily.NormalMeanPrecision | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | in1 | μ(out) :: BayesBase.PointMass | Nothing |
μ(in2) :: ExponentialFamily.NormalMeanPrecision | |||
+ | in1 | μ(out) :: BayesBase.PointMass | Nothing |
μ(in2) :: BayesBase.PointMass | |||
+ | in1 | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
+ | in1 | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(in2) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | in | μ(out) :: Distributions.UnivariateDistribution | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: Distributions.UnivariateDistribution | |||
* | in | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | in | μ(out) :: BayesBase.PointMass | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass | |||
* | in | μ(out) :: ExponentialFamily.GammaDistributionsFamily | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass{<:Real} | |||
* | in | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass{<:AbstractMatrix} | |||
* | in | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass{<:AbstractVector} | |||
* | in | μ(out) :: F | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass{<:Real} | |||
* | in | μ(out) :: ExponentialFamily.MvNormalMeanCovariance | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass{<:AbstractMatrix} | |||
* | in | μ(out) :: ExponentialFamily.MvNormalMeanCovariance | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(A) :: BayesBase.PointMass{<:AbstractVector} | |||
* | in | μ(A) :: BayesBase.PointMass{<:LinearAlgebra.UniformScaling} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(out) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
* | in | μ(A) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(out) :: BayesBase.PointMass{<:Real} | |||
* | in | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(out) :: ExponentialFamily.MvNormalWeightedMeanPrecision | |||
* | in | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(out) :: ExponentialFamily.MvNormalMeanPrecision | |||
* | in | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(out) :: ExponentialFamily.MvNormalMeanCovariance | |||
* | in | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | out | μ(A) :: Distributions.UnivariateDistribution | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Distributions.UnivariateDistribution | |||
* | out | μ(A) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | out | μ(A) :: BayesBase.PointMass{<:LinearAlgebra.UniformScaling} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
* | out | μ(A) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:Real} | |||
* | out | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: ExponentialFamily.MvNormalWeightedMeanPrecision | |||
* | out | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: ExponentialFamily.MvNormalMeanPrecision | |||
* | out | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: ExponentialFamily.MvNormalMeanCovariance | |||
* | out | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | out | μ(A) :: BayesBase.PointMass | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass | |||
* | out | μ(A) :: BayesBase.PointMass{<:Real} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: ExponentialFamily.GammaDistributionsFamily | |||
* | out | μ(A) :: ExponentialFamily.GammaDistributionsFamily | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:Real} | |||
* | out | μ(A) :: BayesBase.PointMass{<:AbstractMatrix} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: F | |||
* | out | μ(A) :: F | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:AbstractMatrix} | |||
* | out | μ(A) :: BayesBase.PointMass{<:AbstractVector} | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | out | μ(A) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:AbstractVector} | |||
* | A | μ(out) :: Distributions.UnivariateDistribution | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Distributions.UnivariateDistribution | |||
* | A | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
* | A | μ(out) :: BayesBase.PointMass | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass | |||
* | A | μ(out) :: ExponentialFamily.GammaDistributionsFamily | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:Real} | |||
* | A | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:AbstractMatrix} | |||
* | A | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:AbstractVector} | |||
* | A | μ(out) :: F | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:Real} | |||
* | A | μ(out) :: ExponentialFamily.MvNormalMeanCovariance | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:AbstractMatrix} | |||
* | A | μ(out) :: ExponentialFamily.MvNormalMeanCovariance | Union{Nothing, MatrixCorrectionTools.AbstractCorrectionStrategy} |
μ(in) :: BayesBase.PointMass{<:AbstractVector} | |||
Distributions.InverseGamma | out | q(α) :: Any | Nothing |
q(θ) :: Any | |||
Distributions.Gamma | out | q(α) :: Any | Nothing |
q(θ) :: Any | |||
Distributions.InverseGamma | out | μ(α) :: BayesBase.PointMass | Nothing |
μ(θ) :: BayesBase.PointMass | |||
Distributions.Gamma | out | μ(α) :: BayesBase.PointMass | Nothing |
μ(θ) :: BayesBase.PointMass | |||
Flow | in | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | FlowMeta{M, <:Unscented} |
Flow | in | μ(out) :: ExponentialFamily.MvNormalWeightedMeanPrecision | FlowMeta{M, <:Linearization} |
Flow | in | μ(out) :: ExponentialFamily.MvNormalMeanPrecision | FlowMeta{M, <:Linearization} |
Flow | in | μ(out) :: ExponentialFamily.MvNormalMeanCovariance | FlowMeta{M, <:Linearization} |
Flow | out | μ(in) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | FlowMeta{M, <:Unscented} |
Flow | out | μ(in) :: ExponentialFamily.MvNormalWeightedMeanPrecision | FlowMeta{M, <:Linearization} |
Flow | out | μ(in) :: ExponentialFamily.MvNormalMeanPrecision | FlowMeta{M, <:Linearization} |
Flow | out | μ(in) :: ExponentialFamily.MvNormalMeanCovariance | FlowMeta{M, <:Linearization} |
ExponentialFamily.MvNormalMeanPrecision | Λ | q(out_μ) :: Any | Nothing |
ExponentialFamily.MvNormalMeanPrecision | Λ | q(out) :: Any | Nothing |
q(μ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | out | μ(μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(Λ) :: Distributions.Wishart | |||
ExponentialFamily.MvNormalMeanPrecision | out | μ(μ) :: BayesBase.PointMass | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | out | μ(μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | out | q(μ) :: BayesBase.PointMass | Nothing |
q(Λ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanPrecision | out | q(μ) :: Any | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | out | μ(μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(Λ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanPrecision | out | μ(μ) :: BayesBase.PointMass | Nothing |
μ(Λ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanPrecision | μ | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(Λ) :: Distributions.Wishart | |||
ExponentialFamily.MvNormalMeanPrecision | μ | μ(out) :: BayesBase.PointMass | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | μ | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | μ | q(out) :: BayesBase.PointMass | Nothing |
q(Λ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanPrecision | μ | q(out) :: Any | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalMeanPrecision | μ | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(Λ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanPrecision | μ | μ(out) :: BayesBase.PointMass | Nothing |
μ(Λ) :: BayesBase.PointMass | |||
GammaMixture{N} | out | q(switch) :: Any | Nothing |
q(a) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
q(b) :: ExponentialFamily.GammaDistributionsFamily | |||
GammaMixture | b | q(out) :: Any | Nothing |
q(switch) :: Any | |||
q(a) :: Any | |||
GammaMixture | a | q(out) :: Any | Nothing |
q(switch) :: Any | |||
q(b) :: ExponentialFamily.GammaDistributionsFamily | |||
GammaMixture{N} | switch | q(out) :: Any | Nothing |
q(a) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
q(b) :: ExponentialFamily.GammaDistributionsFamily | |||
Probit | out | μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, ProbitMeta} |
Probit | out | μ(in) :: BayesBase.PointMass | Union{Nothing, ProbitMeta} |
NOT | out | μ(in) :: Distributions.Bernoulli | Nothing |
Probit | out | q(in) :: BayesBase.PointMass | Union{Nothing, ProbitMeta} |
BIFMHelper | out | q(in) :: Any | Nothing |
Probit | in | μ(out) :: Union{BayesBase.PointMass, Distributions.Bernoulli} | Union{Nothing, ProbitMeta} |
μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
Probit | in | μ(in) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Union{Nothing, ProbitMeta} |
q(out) :: BayesBase.PointMass | |||
Probit | in | μ(out) :: Union{BayesBase.PointMass, Distributions.Bernoulli} | Union{Nothing, ProbitMeta} |
BIFMHelper | in | μ(out) :: Any | Nothing |
NOT | in | μ(out) :: Distributions.Bernoulli | Nothing |
Probit | in | q(out) :: BayesBase.PointMass | Union{Nothing, ProbitMeta} |
Distributions.Beta | out | q(a) :: BayesBase.PointMass | Nothing |
q(b) :: BayesBase.PointMass | |||
Distributions.Uniform | out | q(a) :: BayesBase.PointMass | Nothing |
q(b) :: BayesBase.PointMass | |||
Distributions.Beta | out | μ(a) :: BayesBase.PointMass | Nothing |
μ(b) :: BayesBase.PointMass | |||
Distributions.Uniform | out | μ(a) :: BayesBase.PointMass | Nothing |
μ(b) :: BayesBase.PointMass | |||
Distributions.Wishart | out | q(ν) :: Any | Nothing |
q(S) :: Any | |||
InverseWishart | out | q(ν) :: Any | Nothing |
q(S) :: Any | |||
Distributions.Wishart | out | μ(ν) :: BayesBase.PointMass | Nothing |
q(S) :: Any | |||
InverseWishart | out | μ(ν) :: BayesBase.PointMass | Nothing |
q(S) :: Any | |||
Distributions.Wishart | out | μ(S) :: BayesBase.PointMass | Nothing |
q(ν) :: Any | |||
InverseWishart | out | μ(S) :: BayesBase.PointMass | Nothing |
q(ν) :: Any | |||
Distributions.Wishart | out | μ(ν) :: BayesBase.PointMass | Nothing |
μ(S) :: BayesBase.PointMass | |||
InverseWishart | out | μ(ν) :: BayesBase.PointMass | Nothing |
μ(S) :: BayesBase.PointMass | |||
Poisson | out | q(l) :: ExponentialFamily.GammaDistributionsFamily | Nothing |
Poisson | out | μ(l) :: BayesBase.PointMass | Nothing |
Poisson | l | q(out) :: Any | Nothing |
Poisson | l | μ(out) :: BayesBase.PointMass | Nothing |
MvNormalMeanScalePrecision | γ | q(out_μ) :: Any | Nothing |
MvNormalMeanScalePrecision | γ | q(out) :: Any | Nothing |
q(μ) :: Any | |||
MvNormalMeanScalePrecision | μ | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(γ) :: Any | |||
MvNormalMeanScalePrecision | μ | q(out) :: Any | Nothing |
q(γ) :: Any | |||
MvNormalMeanScalePrecision | out | μ(μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(γ) :: Any | |||
MvNormalMeanScalePrecision | out | q(μ) :: Any | Nothing |
q(γ) :: Any | |||
Transition | in | q(out) :: BayesBase.PointMass | Nothing |
q(a) :: BayesBase.PointMass | |||
Transition | in | q(out) :: Any | Nothing |
q(a) :: ExponentialFamily.MatrixDirichlet | |||
Transition | in | μ(out) :: Distributions.Categorical{P} where P<:Real | |
q(a) :: BayesBase.PointMass | |||
Transition | in | μ(out) :: Distributions.Categorical{P} where P<:Real | Nothing |
q(a) :: ExponentialFamily.MatrixDirichlet | |||
Transition | in | μ(out) :: Distributions.Categorical{P} where P<:Real | Nothing |
μ(a) :: BayesBase.PointMass | |||
Transition | out | μ(in) :: Distributions.DiscreteNonParametric | |
q(a) :: BayesBase.PointMass | |||
Transition | out | μ(in) :: Distributions.Categorical{P} where P<:Real | Nothing |
q(a) :: Distributions.Distribution{Distributions.Matrixvariate, Distributions.Continuous} | |||
Transition | out | q(in) :: Distributions.Categorical{P} where P<:Real | Nothing |
q(a) :: Any | |||
Transition | out | q(in) :: BayesBase.PointMass | Nothing |
q(a) :: BayesBase.PointMass | |||
Transition | out | μ(in) :: Distributions.Categorical{P} where P<:Real | Nothing |
μ(a) :: BayesBase.PointMass | |||
Transition | a | q(out_in) :: BayesBase.Contingency | Nothing |
Transition | a | q(out) :: Any | Nothing |
q(in) :: Distributions.Categorical{P} where P<:Real | |||
SoftDot | γ | q(y_x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(θ) :: Any | |||
SoftDot | γ | q(y) :: Any | Nothing |
q(θ) :: Any | |||
q(x) :: Any | |||
SoftDot | x | μ(y) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(θ) :: Any | |||
q(γ) :: Any | |||
SoftDot | x | q(y) :: Any | Nothing |
q(θ) :: Any | |||
q(γ) :: Any | |||
SoftDot | y | μ(x) :: Any | Nothing |
q(θ) :: Any | |||
q(γ) :: Any | |||
SoftDot | y | q(θ) :: Any | Nothing |
q(x) :: Any | |||
q(γ) :: Any | |||
SoftDot | θ | q(y_x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(γ) :: Any | |||
SoftDot | θ | q(y) :: Any | Nothing |
q(x) :: Any | |||
q(γ) :: Any | |||
AND | out | μ(in1) :: Distributions.Bernoulli | Nothing |
μ(in2) :: Distributions.Bernoulli | |||
OR | out | μ(in1) :: Distributions.Bernoulli | Nothing |
μ(in2) :: Distributions.Bernoulli | |||
AND | in2 | μ(out) :: Distributions.Bernoulli | Nothing |
μ(in1) :: Distributions.Bernoulli | |||
OR | in2 | μ(out) :: Distributions.Bernoulli | Nothing |
μ(in1) :: Distributions.Bernoulli | |||
AND | in1 | μ(out) :: Distributions.Bernoulli | Nothing |
μ(in2) :: Distributions.Bernoulli | |||
OR | in1 | μ(out) :: Distributions.Bernoulli | Nothing |
μ(in2) :: Distributions.Bernoulli | |||
ExponentialFamily.MvNormalMeanCovariance | out | μ(μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(Σ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | out | μ(μ) :: BayesBase.PointMass | Nothing |
q(Σ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | out | q(μ) :: BayesBase.PointMass | Nothing |
q(Σ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanCovariance | out | q(μ) :: Any | Nothing |
q(Σ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | out | μ(μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(Σ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanCovariance | out | μ(μ) :: BayesBase.PointMass | Nothing |
μ(Σ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanCovariance | Σ | q(out_μ) :: Any | Nothing |
ExponentialFamily.MvNormalMeanCovariance | Σ | q(out) :: Any | Nothing |
q(μ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | μ | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(Σ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | μ | μ(out) :: BayesBase.PointMass | Nothing |
q(Σ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | μ | q(out) :: BayesBase.PointMass | Nothing |
q(Σ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanCovariance | μ | q(out) :: Any | Nothing |
q(Σ) :: Any | |||
ExponentialFamily.MvNormalMeanCovariance | μ | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(Σ) :: BayesBase.PointMass | |||
ExponentialFamily.MvNormalMeanCovariance | μ | μ(out) :: BayesBase.PointMass | Nothing |
μ(Σ) :: BayesBase.PointMass | |||
AR | γ | q(y) :: Any | ARMeta |
q(x) :: Any | |||
q(θ) :: Any | |||
AR | γ | q(y_x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | ARMeta |
q(θ) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
AR | x | q(y) :: Any | ARMeta |
q(θ) :: Any | |||
q(γ) :: Any | |||
AR | x | μ(y) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | ARMeta |
q(θ) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
q(γ) :: Any | |||
AR | y | q(x) :: Any | ARMeta |
q(θ) :: Any | |||
q(γ) :: Any | |||
AR | y | μ(x) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | ARMeta |
q(θ) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
q(γ) :: Any | |||
AR | θ | q(y) :: Any | ARMeta |
q(x) :: Any | |||
q(γ) :: Any | |||
AR | θ | q(y_x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | ARMeta |
q(γ) :: Any | |||
Distributions.Bernoulli | out | q(p) :: Any | Nothing |
Distributions.Bernoulli | out | q(p) :: BayesBase.PointMass | Nothing |
Distributions.Categorical{P} where P<:Real | out | q(p) :: BayesBase.PointMass | Nothing |
Distributions.Categorical{P} where P<:Real | out | q(p) :: Distributions.Dirichlet | Nothing |
Distributions.Bernoulli | out | μ(p) :: BayesBase.PointMass | Nothing |
Distributions.Bernoulli | out | μ(p) :: Distributions.Beta | Nothing |
Distributions.Categorical{P} where P<:Real | out | μ(p) :: BayesBase.PointMass | Nothing |
Distributions.Categorical{P} where P<:Real | out | μ(p) :: Distributions.Dirichlet | Nothing |
Distributions.Bernoulli | p | q(out) :: Distributions.Categorical{P} where P<:Real | Nothing |
Distributions.Bernoulli | p | q(out) :: Distributions.Bernoulli | Nothing |
Distributions.Bernoulli | p | q(out) :: BayesBase.PointMass | Nothing |
Distributions.Categorical{P} where P<:Real | p | q(out) :: BayesBase.PointMass | Nothing |
Distributions.Categorical{P} where P<:Real | p | q(out) :: Any | Nothing |
Distributions.Bernoulli | p | μ(out) :: BayesBase.PointMass | Nothing |
DeltaFn | in | μ(in) :: Any | DeltaMeta{M} |
q(ins) :: BayesBase.FactorizedJoint | |||
DeltaFn | in | μ(in) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M, Nothing} |
q(ins) :: ExponentialFamily.JointNormal | |||
DeltaFn | in | μ(in) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M, I} |
q(ins) :: ExponentialFamily.JointNormal | |||
DeltaFn | in | μ(out) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M, I} |
μ(ins) :: Nothing | |||
DeltaFn | in | μ(out) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M, I} |
μ(ins) :: Nothing | |||
DeltaFn | in | μ(out) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M, I} |
μ(ins) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
DeltaFn | in | μ(out) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M, I} |
μ(ins) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | |||
DeltaFn | out | q(ins) :: BayesBase.FactorizedJoint{P} | DeltaMeta{M} |
DeltaFn | out | q(ins) :: BayesBase.FactorizedJoint | DeltaMeta{M} |
DeltaFn | out | μ(ins) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M} |
DeltaFn | out | μ(ins) :: Union{ExponentialFamily.NormalDistributionsFamily{T}, ExponentialFamily.GaussianDistributionsFamily{T}} where T | DeltaMeta{M} |
ExponentialFamily.NormalMeanVariance | v | q(out_μ) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | Nothing |
ExponentialFamily.NormalMeanVariance | v | q(out) :: Any | Nothing |
q(μ) :: Any | |||
ExponentialFamily.NormalMeanVariance | v | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
ExponentialFamily.NormalMeanVariance | v | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(μ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | v | μ(out) :: BayesBase.PointMass | Nothing |
μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | |||
ExponentialFamily.NormalMeanVariance | out | μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | out | μ(μ) :: BayesBase.PointMass | Nothing |
q(v) :: Any | |||
ExponentialFamily.NormalMeanVariance | out | μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(v) :: Any | |||
ExponentialFamily.NormalMeanVariance | out | q(μ) :: BayesBase.PointMass | Nothing |
q(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | out | q(μ) :: Any | Nothing |
q(v) :: Any | |||
ExponentialFamily.NormalMeanVariance | out | μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | out | μ(μ) :: BayesBase.PointMass | Nothing |
μ(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | μ | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(v) :: Any | |||
ExponentialFamily.NormalMeanVariance | μ | μ(out) :: BayesBase.PointMass | Nothing |
q(v) :: Any | |||
ExponentialFamily.NormalMeanVariance | μ | μ(out) :: ReactiveMP.ExponentialLinearQuadratic | |
q(v) :: Any | |||
ExponentialFamily.NormalMeanVariance | μ | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | μ | μ(out) :: BayesBase.PointMass | Nothing |
μ(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | μ | μ(out) :: ReactiveMP.ExponentialLinearQuadratic | |
μ(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | μ | q(out) :: BayesBase.PointMass | Nothing |
q(v) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanVariance | μ | q(out) :: Any | Nothing |
q(v) :: Any | |||
Mixture | switch | μ(out) :: Any | Nothing |
μ(inputs) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
Mixture | out | μ(inputs) :: ManyOf{<:Tuple{Vararg{Any, N}}} | Nothing |
q(switch) :: BayesBase.PointMass | |||
Mixture | out | μ(switch) :: Any | Nothing |
μ(inputs) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
Mixture | inputs | μ(out) :: Any | Nothing |
q(switch) :: BayesBase.PointMass | |||
Mixture | inputs | μ(out) :: Any | Nothing |
μ(switch) :: Any | |||
HalfNormal | out | q(v) :: BayesBase.PointMass | Nothing |
NormalMixture{N} | out | q(switch) :: Any | Nothing |
q(m) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
q(p) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
NormalMixture | p | q(out) :: Any | Nothing |
q(switch) :: Any | |||
q(m) :: Any | |||
NormalMixture | m | q(out) :: Any | Nothing |
q(switch) :: Any | |||
q(p) :: Any | |||
NormalMixture{N} | switch | q(out) :: Any | Nothing |
q(m) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
q(p) :: ManyOf{<:Tuple{Vararg{Any, N}}} | |||
BIFM | zprev | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | BIFMMeta |
μ(in) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
μ(znext) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
BIFM | znext | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | BIFMMeta |
μ(in) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
μ(zprev) :: BayesBase.TerminalProdArgument{<:Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T} | |||
BIFM | out | μ(in) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | BIFMMeta |
μ(zprev) :: BayesBase.TerminalProdArgument{<:Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T} | |||
μ(znext) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
BIFM | in | μ(out) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | BIFMMeta |
μ(zprev) :: BayesBase.TerminalProdArgument{<:Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T} | |||
μ(znext) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
ExponentialFamily.MatrixDirichlet | out | q(a) :: BayesBase.PointMass | Nothing |
Distributions.Dirichlet | out | q(a) :: BayesBase.PointMass{<:AbstractVector} | Nothing |
ExponentialFamily.MatrixDirichlet | out | μ(a) :: BayesBase.PointMass | Nothing |
Distributions.Dirichlet | out | μ(a) :: BayesBase.PointMass{<:AbstractVector} | Nothing |
Distributions.Uniform | out | μ(b) :: BayesBase.PointMass | Nothing |
q(a) :: BayesBase.PointMass | |||
Distributions.Uniform | out | μ(a) :: BayesBase.PointMass | Nothing |
q(b) :: BayesBase.PointMass | |||
ContinuousTransition | W | q(y) :: Any | ContinuousTransitionMeta |
q(x) :: Any | |||
q(a) :: Any | |||
ContinuousTransition | W | q(y_x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | ContinuousTransitionMeta |
q(a) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
ContinuousTransition | a | q(y) :: Any | ContinuousTransitionMeta |
q(x) :: Any | |||
q(a) :: Any | |||
q(W) :: Any | |||
ContinuousTransition | a | q(y_x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | ContinuousTransitionMeta |
q(a) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
q(W) :: Any | |||
ContinuousTransition | x | q(y) :: Any | ContinuousTransitionMeta |
q(a) :: Any | |||
q(W) :: Any | |||
ContinuousTransition | x | μ(y) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | ContinuousTransitionMeta |
q(a) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
q(W) :: Any | |||
ContinuousTransition | y | q(x) :: Any | ContinuousTransitionMeta |
q(a) :: Any | |||
q(W) :: Any | |||
ContinuousTransition | y | μ(x) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | ContinuousTransitionMeta |
q(a) :: Union{ExponentialFamily.MultivariateNormalDistributionsFamily{T}, ExponentialFamily.MultivariateGaussianDistributionsFamily{T}} where T | |||
q(W) :: Any | |||
ExponentialFamily.NormalMeanPrecision | out | μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(τ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | out | μ(μ) :: BayesBase.PointMass | Nothing |
q(τ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | out | q(μ) :: BayesBase.PointMass | Nothing |
q(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | out | q(μ) :: Any | Nothing |
q(τ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | out | μ(μ) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | out | μ(μ) :: BayesBase.PointMass | Nothing |
μ(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | τ | q(out_μ) :: Any | Nothing |
ExponentialFamily.NormalMeanPrecision | τ | q(out) :: Any | Nothing |
q(μ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | μ | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
q(τ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | μ | μ(out) :: BayesBase.PointMass | Nothing |
q(τ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | μ | μ(out) :: ReactiveMP.ExponentialLinearQuadratic | |
q(τ) :: Any | |||
ExponentialFamily.NormalMeanPrecision | μ | μ(out) :: Union{ExponentialFamily.UnivariateNormalDistributionsFamily{T}, ExponentialFamily.UnivariateGaussianDistributionsFamily{T}} where T | Nothing |
μ(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | μ | μ(out) :: BayesBase.PointMass | Nothing |
μ(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | μ | μ(out) :: ReactiveMP.ExponentialLinearQuadratic | |
μ(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | μ | q(out) :: BayesBase.PointMass | Nothing |
q(τ) :: BayesBase.PointMass | |||
ExponentialFamily.NormalMeanPrecision | μ | q(out) :: Any | Nothing |
q(τ) :: Any | |||
ExponentialFamily.MvNormalWeightedMeanPrecision | out | q(ξ) :: Any | Nothing |
q(Λ) :: Any | |||
ExponentialFamily.MvNormalWeightedMeanPrecision | out | μ(ξ) :: BayesBase.PointMass | Nothing |
μ(Λ) :: BayesBase.PointMass | |||
GCV | ω | q(y) :: Any | GCVMetadata |
q(x) :: Any | |||
q(z) :: Any | |||
q(κ) :: Any | |||
GCV | ω | q(y_x) :: Any | GCVMetadata |
q(z) :: Any | |||
q(κ) :: Any | |||
GCV | x | q(y) :: Any | Union{Nothing, GCVMetadata} |
q(z) :: Any | |||
q(κ) :: Any | |||
q(ω) :: Any | |||
GCV | x | μ(y) :: Union{ReactiveMP.ExponentialLinearQuadratic, ExponentialFamily.UnivariateNormalDistributionsFamily, ExponentialFamily.UnivariateGaussianDistributionsFamily} | Union{Nothing, GCVMetadata} |
q(z) :: Any | |||
q(κ) :: Any | |||
q(ω) :: Any | |||
GCV | y | q(x) :: Any | Union{Nothing, GCVMetadata} |
q(z) :: Any | |||
q(κ) :: Any | |||
q(ω) :: Any | |||
GCV | y | μ(x) :: Union{ReactiveMP.ExponentialLinearQuadratic, ExponentialFamily.UnivariateNormalDistributionsFamily, ExponentialFamily.UnivariateGaussianDistributionsFamily} | Union{Nothing, GCVMetadata} |
q(z) :: Any | |||
q(κ) :: Any | |||
q(ω) :: Any | |||
GCV | κ | q(y) :: Any | GCVMetadata |
q(x) :: Any | |||
q(z) :: Any | |||
q(ω) :: Any | |||
GCV | κ | q(y_x) :: Any | GCVMetadata |
q(z) :: Any | |||
q(ω) :: Any | |||
GCV | z | q(y) :: Any | GCVMetadata |
q(x) :: Any | |||
q(κ) :: Any | |||
q(ω) :: Any | |||
GCV | z | q(y_x) :: Any | GCVMetadata |
q(κ) :: Any | |||
q(ω) :: Any | |||
ExponentialFamily.GammaShapeRate | β | q(out) :: Any | Nothing |
q(α) :: Any | |||
ExponentialFamily.GammaShapeRate | α | q(out) :: Any | Nothing |
q(β) :: ExponentialFamily.GammaDistributionsFamily | |||
ExponentialFamily.GammaShapeRate | out | q(α) :: Any | Nothing |
q(β) :: Any | |||
ExponentialFamily.GammaShapeRate | out | μ(α) :: BayesBase.PointMass | Nothing |
μ(β) :: BayesBase.PointMass | |||
out | μ(in1) :: Distributions.Bernoulli | Nothing | |
μ(in2) :: Distributions.Bernoulli | |||
in2 | μ(out) :: Distributions.Bernoulli | Nothing | |
μ(in1) :: Distributions.Bernoulli | |||
in1 | μ(out) :: Distributions.Bernoulli | Nothing | |
μ(in2) :: Distributions.Bernoulli |