Skip to content

Latest commit

 

History

History
125 lines (113 loc) · 5.13 KB

sample_codes.md

File metadata and controls

125 lines (113 loc) · 5.13 KB
Raw

Sample codes

Sample code for simulation and evaluation rumor diffusion in the network

Presented simulation is performed for:

  • two diffusion model - SI with beta = 0.88 and SIS with beta = 0.88 and lambda = 0.005
  • the experiment will be run under Watts-Strogatz topology with n=100, k = 4 and p =0.5
  • there will be 5 randomly selected sources
  • for each diffusion model tehre will be 10 experiments
  • as a result the average number of required iterations to cover all network is presented

The config described above is stored in this file.

from rpasdt.algorithm.models import DiffusionModelSimulationConfig,
    DiffusionSimulationConfig, NetworkSourceSelectionConfig
from rpasdt.algorithm.simulation import perform_diffusion_simulation
from rpasdt.algorithm.taxonomies import DiffusionTypeEnum, NodeStatusEnum,
    GraphTypeEnum, SourceSelectionOptionEnum

result = perform_diffusion_simulation(DiffusionSimulationConfig(
    number_of_experiments=10,
    graph_type=GraphTypeEnum.WATTS_STROGATZ,
    graph_type_properties={"n": 100, "k": 4, "p": 0.5},
    source_selection_config=NetworkSourceSelectionConfig(number_of_sources=5,
                                                         algorithm=SourceSelectionOptionEnum.RANDOM),
    diffusion_models=[
        DiffusionModelSimulationConfig(
            diffusion_model_type=DiffusionTypeEnum.SI,
            diffusion_model_params={"beta": 0.088}
        ),
        DiffusionModelSimulationConfig(
            diffusion_model_type=DiffusionTypeEnum.SIS,
            diffusion_model_params={"beta": 0.088, "lambda": 0.005}
        )
    ]
))
print(result.avg_iteration_to_status_in_population(NodeStatusEnum.INFECTED))

The sample result of the above code:

{'SI': 31.4, 'SIS': 40.4}

Sample code for simulation and evaluation for source detection problem

Presented simulation is performed for:

  • one diffusion model - SI with beta ~ 0.88
  • random source selection model with the constant number of selected sources - 5
  • sources will be evaluated after 20 simulation iterations
  • the source detection will be evaluated with two algorithms: Netlseuth and Rumor center
  • then evaluation of the mentioned algorithms will be presented

The config described above is stored in this file.

from rpasdt.algorithm.models import DiffusionModelSimulationConfig, SourceDetectionSimulationConfig, NetworkSourceSelectionConfig, SourceDetectorSimulationConfig, CommunitiesBasedSourceDetectionConfig
from rpasdt.algorithm.simulation import perform_source_detection_simulation
from rpasdt.algorithm.taxonomies import DiffusionTypeEnum, SourceDetectionAlgorithm

result = perform_source_detection_simulation(SourceDetectionSimulationConfig(

    diffusion_models=[DiffusionModelSimulationConfig(
        diffusion_model_type=DiffusionTypeEnum.SI,
        diffusion_model_params={"beta": 0.08784399402913001}
    )],
    iteration_bunch=20,
    source_selection_config=NetworkSourceSelectionConfig(
        number_of_sources=5,
    ),
    source_detectors={
        "NETLSEUTH": SourceDetectorSimulationConfig(
            alg=SourceDetectionAlgorithm.NET_SLEUTH,
            config=CommunitiesBasedSourceDetectionConfig()
        ),
        "RUMOR_CENTER": SourceDetectorSimulationConfig(
            alg=SourceDetectionAlgorithm.RUMOR_CENTER,
            config=CommunitiesBasedSourceDetectionConfig()
        )
    }
))
print(result.aggregated_results)
for name, results in result.raw_results.items():
    for mm_r in results:
        print(
            f'{name}-{mm_r.real_sources}-{mm_r.detected_sources}-{mm_r.TP}:{mm_r.FN}:{mm_r.FP}')

The sample result of the above code:

{
    'NETLSEUTH': ExperimentSourceDetectionEvaluation(
        avg_error_distance=6.4,
        recall=0.2,
        precision=0.2,
        f1score=0.20000000000000004
    ),
    'RUMOR_CENTER': ExperimentSourceDetectionEvaluation(
        avg_error_distance=4.4,
        recall=0.16,
        precision=0.16,
        f1score=0.16
     )
}
NETLSEUTH-[28, 14, 15, 10, 9]-[23, 3, 24, 10, 9]-2:3:3
NETLSEUTH-[7, 19, 12, 20, 11]-[22, 3, 10, 31, 11]-1:4:4
NETLSEUTH-[15, 11, 26, 8, 0]-[18, 3, 10, 31, 9]-0:5:5
NETLSEUTH-[32, 0, 16, 33, 8]-[23, 3, 24, 10, 9]-0:5:5
NETLSEUTH-[3, 7, 16, 7, 25]-[22, 3, 24, 10, 9]-1:4:4
NETLSEUTH-[10, 29, 25, 12, 18]-[23, 3, 10, 31, 11]-1:4:4
NETLSEUTH-[8, 24, 22, 19, 10]-[23, 3, 24, 10, 9]-2:3:3
NETLSEUTH-[33, 20, 28, 10, 13]-[23, 3, 24, 10, 9]-1:4:4
NETLSEUTH-[9, 14, 25, 22, 7]-[23, 12, 24, 4, 9]-1:4:4
NETLSEUTH-[19, 13, 12, 27, 17]-[23, 13, 24, 9, 11]-1:4:4
RUMOR_CENTER-[28, 14, 15, 10, 9]-[33, 0, 24, 6, 9]-1:4:4
RUMOR_CENTER-[7, 19, 12, 20, 11]-[33, 0, 6, 24, 11]-1:4:4
RUMOR_CENTER-[15, 11, 26, 8, 0]-[33, 0, 5, 28, 9]-1:4:4
RUMOR_CENTER-[32, 0, 16, 33, 8]-[33, 0, 31, 5, 9]-2:3:3
RUMOR_CENTER-[3, 7, 16, 7, 25]-[33, 0, 24, 6, 9]-0:5:5
RUMOR_CENTER-[10, 29, 25, 12, 18]-[33, 0, 6, 31, 11]-0:5:5
RUMOR_CENTER-[8, 24, 22, 19, 10]-[33, 0, 24, 6, 9]-1:4:4
RUMOR_CENTER-[33, 20, 28, 10, 13]-[33, 0, 31, 5, 9]-1:4:4
RUMOR_CENTER-[9, 14, 25, 22, 7]-[33, 0, 24, 10, 9]-1:4:4
RUMOR_CENTER-[19, 13, 12, 27, 17]-[33, 0, 31, 9, 11]-0:5:5