scicom.utilities.statistics
Prune a network.
Classes
Create statistics for communication networks by deletion. |
Functions
|
Generate pruned networks from input. |
Module Contents
- class scicom.utilities.statistics.PruneNetwork(dataframe)
Create statistics for communication networks by deletion.
For a given dataset with sender and receiver information, create a weighted network with igraph. For a given number of iterations, deletion amounts, and deletion types, the algorithm then generates network statistics for randomly sampled subnetworks.
- Parameters:
dataframe (pandas.DataFrame)
- inputDF
- makeNet(dataframe)
Create network from dataframe.
Assumes the existence of sender, receiver and step column names.
- Parameters:
dataframe (pandas.DataFrame)
- Return type:
igraph.Graph
- setSurvivalProb(graph, *, method='agents', ranked=True)
Generate probabilities for different survival modes.
- Parameters:
graph (igraph.Graph)
method (str)
ranked (bool)
- Return type:
pandas.DataFrame
- scaleSurvivalProb(probabilities, *, method='agents')
Scale survival for methods agents and regions.
- Parameters:
probabilities (pandas.DataFrame)
method (str)
- Return type:
pandas.DataFrame
- basicNetStats(graph)
Generate base statistics of network.
- Parameters:
graph (igraph.Graph)
- Return type:
pandas.DataFrame
- netStats(G)
Generate network statistics.
- Parameters:
G (igraph.Graph)
- Return type:
pandas.DataFrame
- deleteFromNetwork(iterations=10, delAmounts=(0.1, 0.25, 0.5, 0.75, 0.9), delTypes=('unif', 'log_normal1', 'exp', 'beta', 'log_normal2', 'log_normal3'), delMethod=('agents', 'regions', 'time'), rankedVals=(True, False))
Run the deletion by sampling.
- Parameters:
iterations (int)
delAmounts (tuple)
delTypes (tuple)
delMethod (tuple)
rankedVals (tuple)
- Return type:
pandas.DataFrame
- scicom.utilities.statistics.prune(modelparameters, network, columns, iterations=10, delAmounts=(0.1, 0.25, 0.5, 0.75, 0.9), delTypes=('unif', 'log_normal1', 'exp', 'beta', 'log_normal2', 'log_normal3'), delMethod=('agents', 'regions', 'time'), rankedVals=(True, False))
Generate pruned networks from input.
Assumes existence of columns “sender”, “receiver”, “sender_location”, “receiver_location” and “step”.
- Parameters:
modelparameters (dict)
network (tuple)
columns (list)
iterations (int)
delAmounts (tuple)
delTypes (tuple)
delMethod (tuple)
rankedVals (tuple)
- Return type:
pandas.DataFrame