Inference problem handed to infer_variational. More...

#include <line/api/infer/infer_variational.h>

Collaboration diagram for line::infer::VariationalSpec< T >:

Static Public Member Functions
static T	unobserved ()
	Sentinel marking an unobserved entry of obsData.

Public Attributes
std::vector< std::array< std::size_t, 3 > >	arcs
	(T x 3) transitions [i j c]; i==0 external source, j==0 sink.
Matrix< T >	x0
	(M x R) initial queue lengths.
std::vector< int >	sched
	(M) discipline codes.
std::vector< T >	nservers
	(M) number of servers.
std::vector< T >	routeprob
	(T) routing probability of each transition.
std::vector< std::size_t >	arcparam
	(T) index in 1..P of the rate governing the transition, 0 when known.
std::vector< T >	arcrate
	(T) known rate for transitions with arcparam == 0.
std::vector< T >	alpha0
	(P) Gamma prior shapes and rates.
std::vector< T >	beta0
std::vector< T >	obsTimes
	(K) observation epochs.
Matrix< T >	obsData
	(K x M*R) observed queue lengths; an unobserved entry is unobserved.
std::vector< T >	obsRange
	(M*R) support size of the uniform contamination.
T	epsilon
	probability that a reading is faulty.
std::vector< T >	capacity
	(M*R) upper bound on the queue length, empty for none.

Detailed Description

template<class T>
struct line::infer::VariationalSpec< T >

Inference problem handed to infer_variational.

Station-class pairs are flattened column-major, so that pair (m,r) sits at index r*M+m, matching the MATLAB, Java and Python specifications. Station and class indices inside arcs are one-based; index 0 marks the external source or sink.

Definition at line 74 of file infer_variational.h.

Member Function Documentation

◆ unobserved()

template<class T>

T line::infer::VariationalSpec< T >::unobserved ( )

inlinestatic

Sentinel marking an unobserved entry of obsData.

Definition at line 108 of file infer_variational.h.

Referenced by line::infer::infer_variational().

Member Data Documentation

◆ alpha0

template<class T>

std::vector<T> line::infer::VariationalSpec< T >::alpha0

(P) Gamma prior shapes and rates.

Definition at line 90 of file infer_variational.h.

Referenced by line::infer::infer_variational().

◆ arcparam

template<class T>

std::vector<std::size_t> line::infer::VariationalSpec< T >::arcparam

(T) index in 1..P of the rate governing the transition, 0 when known.

Definition at line 86 of file infer_variational.h.

Referenced by line::infer::infer_variational().

◆ arcrate

template<class T>

std::vector<T> line::infer::VariationalSpec< T >::arcrate

(T) known rate for transitions with arcparam == 0.

Definition at line 88 of file infer_variational.h.

Referenced by line::infer::infer_variational().

◆ arcs

template<class T>

std::vector<std::array<std::size_t, 3> > line::infer::VariationalSpec< T >::arcs

(T x 3) transitions [i j c]; i==0 external source, j==0 sink.

Definition at line 76 of file infer_variational.h.

Referenced by line::infer::infer_variational().

◆ beta0

template<class T>

std::vector<T> line::infer::VariationalSpec< T >::beta0

Definition at line 90 of file infer_variational.h.

Referenced by line::infer::infer_variational().

◆ capacity

template<class T>

std::vector<T> line::infer::VariationalSpec< T >::capacity

(M*R) upper bound on the queue length, empty for none.

In a closed network this is the chain population, and clamping the load there keeps the expanded state space from crediting a station with more jobs than the network holds.

Definition at line 105 of file infer_variational.h.

Referenced by line::infer::infer_variational().