nifty.graph.opt.mincut

Functions

nifty.graph.opt.mincut.mincutObjective()

Classes

class nifty.graph.opt.mincut.MincutBaseMincutObjectiveEdgeContractionGraphUndirectedGraph
optimize()
class nifty.graph.opt.mincut.MincutBaseMincutObjectiveUndirectedGraph
optimize()
class nifty.graph.opt.mincut.MincutCcFusionMoveBasedFactoryMincutObjectiveUndirectedGraph
class nifty.graph.opt.mincut.MincutCcFusionMoveBasedMincutObjectiveUndirectedGraph
class nifty.graph.opt.mincut.MincutCcFusionMoveBasedSettingsTypeMincutObjectiveUndirectedGraph
fusionMoveSettings
numberOfIterations
numberOfThreads
proposalGenerator
stopIfNoImprovement
class nifty.graph.opt.mincut.MincutCcFusionMoveSettingsMincutObjectiveUndirectedGraph
mincutFactory
class nifty.graph.opt.mincut.MincutObjectiveEdgeContractionGraphUndirectedGraph
evalNodeLabels()
graph
static greedyAdditiveFactory(weightStopCond=0.0, nodeNumStopCond=-1.0, improve=True)
static mincutCcFusionMoveBasedFactory(proposalGenerator=None, numberOfThreads=1, numberOfIterations=1000, stopIfNoImprovement=100, fusionMoveSettings=None)

factory function for a cc-fusion move based mincut solver

Parameters:
  • proposalGenerator (None, optional) – Proposal generator (default watershedProposalGenerator)
  • numberOfThreads (int, optional) – (default 1)
  • numberOfIterations (int, optional) – Maximum number of iterations(default 1000)
  • stopIfNoImprovement (int, optional) – Stop after n iterations without improvement (default 100)
  • fusionMoveSettings (FusionMoveSettings, optional) – The settings of the underlaying mincutCcFusionMove
Returns:

MincutCcFusionMoveBasedFactory
Return type:

TYPE
static mincutQpboFactory(improve=True)
static verboseVisitor(visitNth=1, timeLimit=0)
static watershedProposalGenerator(sigma=1.0, numberOfSeeds=0.1, seedingStrategie='SEED_FROM_NEGATIVE')

factory function for a watershed based proposal generator for mincutCcFusionMoveBased

Parameters:
  • sigma (float, optional) – The weights are perturbed by a additive Gaussian noise n(0,sigma) (default 0.0)
  • numberOfSeeds (float, optional) – Number of seed to generate. A number smaller as one will be interpreted as a fraction of the number of nodes (default 0.1)
  • seedingStrategie (str, optional) –

    Can be: - ‘SEED_FROM_NEGATIVE’ : All negative weighted edges

    can be used to generate seeds.
    • ’SEED_FROM_ALL’ : All edges
      can be used to generate seeds.
Returns:

parameter object used construct a WatershedProposalGenerator
Return type:

TYPE
class nifty.graph.opt.mincut.MincutObjectiveUndirectedGraph
evalNodeLabels()
graph
static greedyAdditiveFactory(weightStopCond=0.0, nodeNumStopCond=-1.0, improve=True)
static mincutCcFusionMoveBasedFactory(proposalGenerator=None, numberOfThreads=1, numberOfIterations=1000, stopIfNoImprovement=100, fusionMoveSettings=None)

factory function for a cc-fusion move based mincut solver

Parameters:
  • proposalGenerator (None, optional) – Proposal generator (default watershedProposalGenerator)
  • numberOfThreads (int, optional) – (default 1)
  • numberOfIterations (int, optional) – Maximum number of iterations(default 1000)
  • stopIfNoImprovement (int, optional) – Stop after n iterations without improvement (default 100)
  • fusionMoveSettings (FusionMoveSettings, optional) – The settings of the underlaying mincutCcFusionMove
Returns:

MincutCcFusionMoveBasedFactory
Return type:

TYPE
static mincutQpboFactory(improve=True)
static verboseVisitor(visitNth=1, timeLimit=0)
static watershedProposalGenerator(sigma=1.0, numberOfSeeds=0.1, seedingStrategie='SEED_FROM_NEGATIVE')

factory function for a watershed based proposal generator for mincutCcFusionMoveBased

Parameters:
  • sigma (float, optional) – The weights are perturbed by a additive Gaussian noise n(0,sigma) (default 0.0)
  • numberOfSeeds (float, optional) – Number of seed to generate. A number smaller as one will be interpreted as a fraction of the number of nodes (default 0.1)
  • seedingStrategie (str, optional) –

    Can be: - ‘SEED_FROM_NEGATIVE’ : All negative weighted edges

    can be used to generate seeds.
    • ’SEED_FROM_ALL’ : All edges
      can be used to generate seeds.
Returns:

parameter object used construct a WatershedProposalGenerator
Return type:

TYPE
class nifty.graph.opt.mincut.MincutVisitorBaseMincutObjectiveEdgeContractionGraphUndirectedGraph
class nifty.graph.opt.mincut.MincutVisitorBaseMincutObjectiveUndirectedGraph
class nifty.graph.opt.mincut.SolverFactoryBaseMincutObjectiveEdgeContractionGraphUndirectedGraph
create()
class nifty.graph.opt.mincut.SolverFactoryBaseMincutObjectiveUndirectedGraph
create()
class nifty.graph.opt.mincut.VerboseVisitorMincutObjectiveEdgeContractionGraphUndirectedGraph
runtimeSolver
runtimeTotal
stopOptimize()
timeLimitSolver
timeLimitTotal
class nifty.graph.opt.mincut.VerboseVisitorMincutObjectiveUndirectedGraph
runtimeSolver
runtimeTotal
stopOptimize()
timeLimitSolver
timeLimitTotal
class nifty.graph.opt.mincut.WatershedProposalGeneratorFactoryMincutObjectiveUndirectedGraph