The pygenstability cli

Command line interface.

cli

App initialisation.

Usage

cli [OPTIONS] COMMAND [ARGS]...

plot_communities

Plot communities on networkx graph.

Usage

cli plot_communities [OPTIONS] GRAPH_FILE RESULTS_FILE

Arguments

GRAPH_FILE

Required argument

RESULTS_FILE

Required argument

plot_scan

Plot results in scan plot.

Usage

cli plot_scan [OPTIONS] RESULTS_FILE

Arguments

RESULTS_FILE

Required argument

run

Run pygenstability.

graph_file: path to either a .pkl with adjacency matrix in sparse format, or a text file with three columns encoding node indices and edge weight. The columns need a header, or the first line will be dropped. Notice that doubled edges with opposite orientations are needed for symetric graph.

See https://barahona-research-group.github.io/PyGenStability/ for more information.

Usage

cli run [OPTIONS] GRAPH_FILE

Options

--constructor <constructor>

Name of the quality constructor.

Default:

'linearized'

--min-scale <min_scale>

Minimum scale.

Default:

-2.0

--max-scale <max_scale>

Maximum scale.

Default:

0.5

--n-scale <n_scale>

Number of scale steps.

Default:

20

--log-scale <log_scale>

Use linear or log scales.

Default:

True

--n-tries <n_tries>

Number of Louvain evaluations.

Default:

100

--NVI, --no-NVI

Compute the normalized variation of information between runs.

Default:

True

--n-NVI <n_nvi>

Number of randomly chosen runs to estimate the NVI.

Default:

20

--postprocessing, --no-postprocessing

Apply the final postprocessing step.

Default:

True

--ttprime, --no-ttprime

Compute the ttprime matrix.

Default:

True

--spectral-gap, --no-spectral-gap

Normalize scale by spectral gap.

Default:

True

--result-file <result_file>

Path to the result file.

Default:

'results.pkl'

--n-workers <n_workers>

Number of workers for multiprocessing.

Default:

4

--tqdm-disable <tqdm_disable>

disable progress bars

Default:

False

--method <method>

Method to solve modularity, either Louvain or Leiden

Default:

'louvain'

--with-optimal-scales, --no-with-optimal-scales

Search for optimal scales

Default:

True

--exp-comp-mode <exp_comp_mode>

Method to compute matrix exponential, can be ‘spectral’ or ‘expm’

Default:

'spectral'

--seed <seed>

Seed for the random number generator; omit for non-deterministic runs.

Arguments

GRAPH_FILE

Required argument