cosifer.combiners package¶
Submodules¶
cosifer.combiners.cit module¶
Combiner for interaction tables.
-
class
cosifer.combiners.cit.
CombineInteractionTables
(combine_tables=<function mean_table>, interaction_symbol='<->', name='cit', **kwargs)¶ Bases:
cosifer.combiners.network_combiner.NetworkCombiner
Combine interaction tables representing networks using a function.
-
name
¶ name of the combiner.
- Type
str
-
combine_tables
¶ function to combine interaction tables.
- Type
function
-
interaction_symbol
¶ symbol used to indicate interactions in the index of the dataframe.
- Type
str
-
graph
¶ combined graph.
- Type
cosifer.collections.Graph
-
-
cosifer.combiners.cit.
combine_tables
(table_list, interaction_symbol='<->', processing_concatenated_intensities_fn=<function <lambda>>, reduce_fn=<function <lambda>>, **kwargs)¶ Compute the combined intensities from a list of interaction tables.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
processing_concatenated_intensities_fn (function, optional) – function to apply on the concatenated intensities. Defaults to identity.
reduce_fn (function, optional) – function to reduce intensities over dataframe rows. Defaults to mean.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
get_scaled_ranks
(dataframe)¶ Scaled ranks from an intensity dataframe.
- Parameters
dataframe (pd.DataFrame) – intensity dataframe.
- Returns
scaled ranks dataframe
- Return type
pd.DataFrame
-
cosifer.combiners.cit.
hard_mean_scaled_ranks_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the mean on the scaled ranks without considering interaction existence.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
hard_mean_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the mean on the intensities without considering interaction existence.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
max_scaled_ranks_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the maximum on the scaled ranks.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
max_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the maximum on the intensities.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
mean_scaled_ranks_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the mean on the scaled ranks.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
mean_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the mean on the intensities.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
median_scaled_ranks_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the median on the scaled ranks.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
median_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the median on the intensities.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
min_scaled_ranks_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the minimum on the scaled ranks.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
-
cosifer.combiners.cit.
min_table
(table_list, interaction_symbol='<->', **kwargs)¶ Compute the minimum on the intensities.
- Parameters
table_list (list) – a list of interaction tables.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
the combined interaction table.
- Return type
cosifer.combiners.core module¶
Core combiner utilities.
-
cosifer.combiners.core.
combined_df_to_interaction_table
(combined, table_list, interaction_symbol='<->')¶ Transform combined intensities dataframe into an InteractionTable.
- Parameters
combined (pd.DataFrame) – combined intensities dataframe.
table_list (list) – a list of InteractionTable objects.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
- an InteractionTable representing the combined
intensities.
- Return type
-
cosifer.combiners.core.
concatenate_intensities
(table_list, threshold_rate=None)¶ Concatenate intensities from a list of InteractionTable objects.
- Parameters
table_list (list) – a list of InteractionTable objects.
threshold_rate (float, optional) – threshold rate for the NAs. Defaults to None, a.k.a no threshold applied.
- Returns
- an InteractionTable representing the combined
intensities.
- Return type
cosifer.combiners.network_combiner module¶
Interface for a network combiner.
-
class
cosifer.combiners.network_combiner.
NetworkCombiner
(filepath=None, **kwargs)¶ Bases:
cosifer.handlers.network_handler.NetworkHandler
Abstract interface for a network combiner.
-
trained
¶ flag to indicate whether the combiner is already trained.
- Type
bool
-
combine
(results_list)¶ Apply the combination method. Checking whehter the combiner has been already trained.
- Parameters
results_list (list) – a list of InteractionTable objects.
-
cosifer.combiners.snf module¶
SNF combiner.
-
class
cosifer.combiners.snf.
SNF
(interaction_symbol='<->', name='snf', **kwargs)¶ Bases:
cosifer.combiners.network_combiner.NetworkCombiner
Combine interaction tables representing networks using SNF.
-
interaction_symbol
¶ symbol used to indicate interactions in the index of the dataframe.
- Type
str
-
name
¶ name of the combiner.
- Type
str
-
-
cosifer.combiners.snf.
compute_snf
(results_list, labels, K=20, T=10, snf=rpy2.robjects.packages.Package as a <module 'SNFtool'>)¶ Compute combination via SNF.
- Parameters
results_list (list) – a list of InteractionTable objects.
labels (list) – labels to consider.
K (int, optional) – number of nearest neighbors. Defaults to 20.
T (int, optional) – number of steps in the diffusion process. Defaults to 10.
snf (object, optional) – SNF rpy2 object. Defaults to importr(‘SNFtool’).
- Returns
the combined graph.
- Return type
cosifer.combiners.summa module¶
Combiner using SUMMA.
-
class
cosifer.combiners.summa.
Summa
(interaction_symbol='<->', name='summa', **kwargs)¶ Bases:
cosifer.combiners.network_combiner.NetworkCombiner
SUMMA algorithm for combining results.
-
interaction_symbol
¶ symbol used to indicate interactions in the index of the dataframe.
- Type
str
-
name
¶ name of the combiner.
- Type
str
-
summa_object
¶ SUMMA object.
- Type
pySUMMA.Summa
-
summa_object
= None
-
-
cosifer.combiners.summa.
fill_na_ranks
(a_series)¶ Fill NA ranks in a pd.Series.
- Parameters
a_series (pd.Series) – series to fill.
- Returns
the filled series.
- Return type
pd.Series
-
cosifer.combiners.summa.
summa_scores_table
(table_list, interaction_symbol='<->', **kwargs)¶ Apply SUMMA as a method on list of interaction tables.
- Parameters
table_list (list) – a list of InteractionTable objects.
interaction_symbol (str, optional) – symbol used to indicate interactions in the index of the dataframe. Defaults to ‘<->’.
- Returns
interaction table with SUMMA scores.
- Return type
Module contents¶
Combiner module.