cosifer.inferencers package¶

Submodules¶

cosifer.inferencers.aracne module¶

Aracne inferencer.

class cosifer.inferencers.aracne.Aracne(estimator, disc='none', method='ARACNE', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

Aracne inferencer implementation.

estimator¶

estimator type.

Type: str

disc¶

discretization type.

Type: str

method¶

name of the method.

Type: str

cosifer.inferencers.clr module¶

CLR inferencer.

class cosifer.inferencers.clr.CLR(estimator, disc='none', method='CLR', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

CLR inferencer implementation.

estimator¶

estimator type.

Type: str

disc¶

discretization type.

Type: str

method¶

name of the method.

Type: str

cosifer.inferencers.correlation module¶

Correlation inferencer.

class cosifer.inferencers.correlation.Correlation(method=None, correction=None, confidence_threshold=0.05, **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

Correlation inferencer.

method¶

correlation method.

Type: str

correction¶

correction method.

Type: str

confidence_threshold¶

confidence threshold.

Type: float

method = None

cosifer.inferencers.funchisq module¶

FunChisq inferencer.

class cosifer.inferencers.funchisq.FunChisq(k_min=3, k_max=7, k_step=1, method='FunChisq', correction=None, confidence_threshold=0.05, undirected=True, **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

FunChisq inferencer.

k_min¶

minimum number of quantization bins.

Type: int

k_max¶

maximum number of quantization bins.

Type: int

k_step¶

number of steps for bins search.

Type: int

method¶

name of the method.

Type: str

correction¶

correction method.

Type: str

confidence_threshold¶

confidence threshold.

Type: float

undirected¶

flag to indicate an undirected network.

Type: bool

cosifer.inferencers.funchisq.sort_interaction_entities(row)¶

Sort the entities over a row in lexicographic order.

Parameters

row (pd.Series) – row containing the entities to be sorted.

Returns

a list containing the sorted entities and the rest of the: elements of the row unchanged.

Return type

list

cosifer.inferencers.genie3 module¶

GENIE3 inferencer.

class cosifer.inferencers.genie3.GENIE3(tree_method='RF', k='sqrt', n_trees=1000, regulators=<rpy2.rinterface.NULLType object> [RTYPES.NILSXP], targets=<rpy2.rinterface.NULLType object> [RTYPES.NILSXP], n_cores=4, verbose=False, method='GENIE3', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

GENIE3 inferencer.

tree_method¶

tree method.

Type: str

k¶

k criterion.

Type: str

n_trees¶

number of trees.

Type: int

regulators¶

known regulators.

Type: object

targets¶

known targets.

Type: object

n_cores¶

number of cores.

Type: int

verbose¶

toggle verbosity.

Type: bool

method¶

name of the method.

Type: str

cosifer.inferencers.glasso module¶

Glasso inferencer.

class cosifer.inferencers.glasso.Glasso(correction=None, method='gLasso', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

Glasso inferencer.

method¶

name of the method.

Type: str

cosifer.inferencers.jrf module¶

JRF inferencer.

class cosifer.inferencers.jrf.JointRandomForest(ntree=500, mtry=None, merger=<function JointRandomForest.<lambda>>, correction=None, method='JRF', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

JRF inferencer.

ntree¶

number of trees.

Type: int

mtry¶

number of variables for splitting.

Type: int

merger¶

a merger function.

Type: function

method¶

name of the method.

Type: str

cosifer.inferencers.mrnet module¶

MRNET inferencer.

class cosifer.inferencers.mrnet.MRNET(estimator, disc='none', method='MRNET', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

MRNET inferencer.

estimator¶

estimator type.

Type: str

disc¶

discretization type.

Type: str

method¶

name of the method.

Type: str

cosifer.inferencers.network_inferencer module¶

NetworkInferencer abstract interface.

class cosifer.inferencers.network_inferencer.NetworkInferencer(**kwargs)¶

Bases: cosifer.handlers.network_handler.NetworkHandler

Network inferencer interface.

graph¶

graph representing the network.

Type: cosifer.collections.graph,Graph

trained¶

flag to indicate whether the inference has been performed.

Type: bool

graph = None

infer_network(data)¶

Infer the network.

Parameters: data (pd.DataFrame) – data to be used for the inference.

trained = False

cosifer.inferencers.tigress module¶

TIGRESS inferencer.

class cosifer.inferencers.tigress.TIGRESS(tf_list=<rpy2.rinterface.NULLType object> [RTYPES.NILSXP], k=-1, alpha=0.2, n_steps_lars=5, n_bootstrap=1000, scoring='area', verbose=False, use_parallel=True, n_cores=4, method='TIGRESS', **kwargs)¶

Bases: cosifer.inferencers.network_inferencer.NetworkInferencer

TIGRESS inferencer.

tf_list¶

list of transcription factor.

Type: object

k¶

number of edges to return.

Type: int

alpha¶

alpha parameter.

Type: float

n_step_lars¶

number of LARS steps.

Type: int

n_bootstrap¶

bootstrap number.

Type: int

scoring¶

scoring criterion.

Type: str

verbose¶

toggle verbosity.

Type: bool

use_parallel¶

enable parallelism.

Type: bool

n_cores¶

number of cores.

Type: int

method¶

name of the method.

Type: str

Module contents¶

Inferencer module.