Need help? Have a feature request? Please check out the tethne-users group .

tethne.analyze package

Submodules

tethne.analyze.collection module

Methods for analyzing GraphCollections.

algorithm	Apply a method from NetworkX to all networkx.Graph objects in the GraphCollection G.
attachment_probability	Calculates the attachment probability for each node at each time-step.
connected	Performs analysis methods from networkx.connected on each graph in the collection.
delta
node_global_closeness_centrality

tethne.analyze.collection.algorithm(G, method_name, **kwargs)

Apply a method from NetworkX to all networkx.Graph objects in the GraphCollection G.

For options, see the list of algorithms in the NetworkX documentation. Not all of these have been tested.

Parameters:

G : GraphCollection

The GraphCollection to analyze. The specified method will be applied to each graph in G.

method_name : string

Name of a method in NetworkX to execute on graph collection.

**kwargs

A list of keyword arguments that should correspond to the parameters of the specified method.

Returns:

results : dict

Indexed by element (node or edge) and graph index (e.g. date).

Raises:

ValueError

If no such method exists.

Examples

Betweenness centrality: (G is a GraphCollection)

>>> from tethne.analyze import collection
>>> BC = collection.algorithm(G, 'betweenness_centrality')
>>> print BC[0]
{1999: 0.010101651117889644,
2000: 0.0008689093723107329,
2001: 0.010504898852426189,
2002: 0.009338654511194512,
2003: 0.007519105636349891}

tethne.analyze.collection.attachment_probability(G, raw=False)

Calculates the attachment probability for each node at each time-step.

The attachment probability for a node at a particular graph state t is the probability that the next edge added to the graph will accrue to that node. The MLE for attachment probability is thus the observed fraction of all new edges in graph state t + 1 that accrue to a particular node.

Note that values will only be calculated for nodes present in state t. In other words, if in t + 1 a new node is introduced who also accrues new edges, that node will not appear in the results for state t.

Parameters:

G : GraphCollection

raw : bool

(default: False) If True, nodes are represented by their integer ids in G, rather than their label.

Returns:

probs : dict

Keyed by index in G, and then by node. If raw is True, node keys will be integer indices from the GraphCollection’s node_index.

Examples

>>> from tethne.readers.wos import read
>>> corpus = read("/path/to/my/data")
>>> from tethne import coauthors, GraphCollection
>>> collection = GraphCollection(corpus, coauthors)
>>> from tethne.analyze.collection import attachment_probability
>>> probs = attachment_probability(collection)

tethne.analyze.collection.connected(G, method_name, **kwargs)

Performs analysis methods from networkx.connected on each graph in the collection.

Parameters:

G : GraphCollection

The GraphCollection to analyze. The specified method will be applied to each graph in G.

method : string

Name of method in networkx.connected.

**kwargs : kwargs

Keyword arguments, passed directly to method.

Returns:

results : dict

Keys are graph indices, values are output of method for that graph.

Raises:

ValueError

If name is not in networkx.connected, or if no such method exists.

tethne.analyze.corpus module

Methods for analyzing Corpus objects.

burstness	Estimate burstness profile for the topn features (or flist) in feature.
feature_burstness	Estimate burstness profile for a feature over the 'date' axis.
sigma	Calculate sigma (from Chen 2009) for all of the nodes in a GraphCollection.

tethne.analyze.corpus.burstness(corpus, featureset_name, features=[], k=5, topn=20, perslice=False, normalize=True, **kwargs)

Estimate burstness profile for the topn features (or flist) in feature.

Uses the popular burstness automaton model inroduced by Kleinberg (2002).

Parameters:

corpus : Corpus

feature : str

Name of featureset in corpus. E.g. 'citations'.

k : int

(default: 5) Number of burst states.

topn : int or float {0.-1.}

(default: 20) Number (int) or percentage (float) of top-occurring features to return. If flist is provided, this parameter is ignored.

perslice : bool

(default: False) If True, loads topn features per slice. Otherwise, loads topn features overall. If flist is provided, this parameter is ignored.

flist : list

List of features. If provided, topn and perslice are ignored.

normalize : bool

(default: True) If True, burstness is expressed relative to the hightest possible state (k-1). Otherwise, states themselves are returned.

kwargs : kwargs

Parameters for burstness automaton HMM.

Returns:

B : dict

Keys are features, values are tuples of ( dates, burstness )

Examples

>>> from tethne.analyze.corpus import burstness
>>> B = burstness(corpus, 'abstractTerms', flist=['process', 'method']
>>> B['process']
([1990, 1991, 1992, 1993], [0., 0.4, 0.6, 0.])

tethne.analyze.corpus.feature_burstness(corpus, featureset_name, feature, k=5, normalize=True, s=1.1, gamma=1.0, **slice_kwargs)

Estimate burstness profile for a feature over the 'date' axis.

Parameters:

corpus : Corpus

feature : str

Name of featureset in corpus. E.g. 'citations'.

findex : int

Index of feature in corpus.

k : int

(default: 5) Number of burst states.

normalize : bool

(default: True) If True, burstness is expressed relative to the hightest possible state (k-1). Otherwise, states themselves are returned.

kwargs : kwargs

Parameters for burstness automaton HMM.

tethne.analyze.corpus.sigma(G, corpus, featureset_name, B=None, **kwargs)

Calculate sigma (from Chen 2009) for all of the nodes in a GraphCollection.

You can set parameters for burstness estimation using kwargs:

Parameter	Description
s	Scaling parameter ( > 1.)that controls graininess of burst detection. Lower values make the model more sensitive. Defaults to 1.1.
gamma	Parameter that controls the ‘cost’ of higher burst states. Defaults to 1.0.
k	Number of burst states. Defaults to 5.

Parameters:

G : GraphCollection

corpus : Corpus

feature : str

Name of a featureset in corpus.

Examples

Assuming that you have a Corpus generated from WoS data that has been sliced by date.

>>> # Generate a co-citation graph collection.
>>> from tethne import GraphCollection
>>> kwargs = { 'threshold':2, 'topn':100 }
>>> G = GraphCollection()
>>> G.build(corpus, 'date', 'papers', 'cocitation', method_kwargs=kwargs)

>>> # Calculate sigma. This may take several minutes, depending on the
>>> #  size of your co-citaiton graph collection.
>>> from tethne.analyze.corpus import sigma
>>> G = sigma(G, corpus, 'citations')

>>> # Visualize...
>>> from tethne.writers import collection
>>> collection.to_dxgmml(G, '~/cocitation.xgmml')

In the visualization below, node and label sizes are mapped to sigma, and border width is mapped to citations.

tethne.analyze.features module

Methods for analyzing featuresets.

cosine_distance
cosine_similarity	Calculate cosine similarity for sparse feature vectors.
distance
kl_divergence	Calculate Kullback-Leibler distance.

tethne.analyze.features.angular_similarity(F_a, F_b)

Calculate the angular similarity for sparse feature vectors.

Unlike cosine_similarity, this is a true distance metric.

Parameters:

F_a : Feature

F_b : Feature

Returns:

similarity : float

Cosine similarity.

tethne.analyze.features.cosine_similarity(F_a, F_b)

Calculate cosine similarity for sparse feature vectors.

Parameters:

F_a : Feature

F_b : Feature

Returns:

similarity : float

Cosine similarity.

tethne.analyze.features.kl_divergence(V_a, V_b)

Calculate Kullback-Leibler distance.

Uses the smoothing method described in Bigi 2003 to facilitate better comparisons between vectors describing wordcounts.

Parameters:

V_a : list

V_b : list

Returns:

divergence : float

KL divergence.

tethne.analyze.graph module

Methods for network analysis.

global_closeness_centrality

Calculates global closeness centrality for one or all nodes in the network.

tethne.analyze.graph.global_closeness_centrality(g, node=None, normalize=True)

Calculates global closeness centrality for one or all nodes in the network.

See node_global_closeness_centrality() for more information.

Parameters:

g : networkx.Graph

normalize : boolean

If True, normalizes centrality based on the average shortest path length. Default is True.

Returns:

C : dict

Dictionary of results, with node identifiers as keys and gcc as values.

Module contents

Methods for analyzing Corpus, GraphCollection, and networkx.Graph objects.

collection	Methods for analyzing GraphCollections.
corpus	Methods for analyzing Corpus objects.
features	Methods for analyzing featuresets.
graph	Methods for network analysis.