The result of a JMT solve: the shared AvgResult plus what only JMT reports. More...

#include <line/solvers/wrappers/jmt/solver_jmt.h>

Collaboration diagram for line::jmt::JmtResult< T >:

Public Attributes
mva::AvgResult< T >	avg
Matrix< T >	QCI
Matrix< T >	UCI
Matrix< T >	RCI
Matrix< T >	TCI
Matrix< T >	ACI
	confidence half-widths, empty when disabled
Matrix< T >	Weight
Matrix< T >	MemOcc
	FCR-only: weighted and memory occupation.
Matrix< T >	TNfcr
Matrix< T >	DropRateNfcr
	(nregions x nclasses) carried and lost rate
std::map< std::size_t, std::vector< T > >	cache_hit_prob
	Per Cache node (1-based node index), the per-class hit probability.
T	log_norm_const = num_traits<T>::from_double(std::numeric_limits<double>::quiet_NaN())
	result.Prob.logNormConstAggr, the log normalizing constant JMVA reports.

Detailed Description

template<class T>
struct line::jmt::JmtResult< T >

The result of a JMT solve: the shared AvgResult plus what only JMT reports.

The FCR rows extend the metric matrices past nstations, which is the reference's layout (getResults.m allocates nstations + nregions rows) and is what makes a region's measures reachable through the same table as a station's.

Definition at line 493 of file solver_jmt.h.

Member Data Documentation

◆ ACI

template<class T>

Matrix<T> line::jmt::JmtResult< T >::ACI

confidence half-widths, empty when disabled

Definition at line 495 of file solver_jmt.h.

Referenced by line::jmt::jmt_map_measures().

◆ avg

template<class T>

mva::AvgResult<T> line::jmt::JmtResult< T >::avg

Definition at line 494 of file solver_jmt.h.

Referenced by line::NetworkSolver::avg_table(), line::jmt::jmt_get_cdf_resp_t(), line::jmt::jmt_map_measures(), line::jmt::jmt_parse_jmva(), line::jmt::jmt_region_losses(), and line::jmt::solver_jmt_run_analyzer().

◆ cache_hit_prob

template<class T>

std::map<std::size_t, std::vector<T> > line::jmt::JmtResult< T >::cache_hit_prob

Per Cache node (1-based node index), the per-class hit probability.

Definition at line 499 of file solver_jmt.h.

Referenced by line::jmt::jmt_map_measures().

◆ DropRateNfcr

template<class T>

Matrix<T> line::jmt::JmtResult< T >::DropRateNfcr

(nregions x nclasses) carried and lost rate

Definition at line 497 of file solver_jmt.h.

Referenced by line::jmt::jmt_region_losses().

◆ log_norm_const

template<class T>

T line::jmt::JmtResult< T >::log_norm_const = num_traits<T>::from_double(std::numeric_limits<double>::quiet_NaN())

result.Prob.logNormConstAggr, the log normalizing constant JMVA reports.

NaN on the simulation path and on any JMVA algorithm that does not compute one – an AMVA approximation has no G to report – rather than zero, which is a legitimate value of a log constant.

Definition at line 507 of file solver_jmt.h.

Referenced by line::jmt::jmt_parse_jmva().