GraSPy 0.3

Announcement: GraSPy 0.3

We're happy to announce the release of GraSPy 0.3! GraSPy is a Python package for
understanding the properties of random graphs that arise from modern datasets, such as
social networks and brain networks.

For more information, please visit our website and our tutorials

Highlights

This release is the result of over 5 months of work with over 11 pull requests by
7 contributors. Highlights include:

• Added seeded graph matching as a capability for graph matching, renamed graph matching class to GraphMatch
• Added functions for simulating a pair of correlated RDPG graphs.
• Deprecated Python 3.5
• Added different backend hypothesis tests for the LatentDistributionTest from Hyppo
• Added a correction to make LatentDistributionTest valid for differently sized graphs

Improvements

• Updated default value of rescale in RDPG simulation
• Updated default value of scaled in MASE estimation
• Improved error throwing in AutoGMM
• Clarified the API for inference submodule

API Changes

• FastApproximateQAP was renamed to GraphMatch
• fit method of LatentDistributionTest and LatentPositionTest now returns self instead of a p-value

Deprecations

• Python 3.5

Contributors to this release

• Jaewon Chung
• Benjamin Pedigo
• Ali Saad-Eldin
• Shan Qiu
• Bijan Varjavand
• Anton Alyakin (new contributor!)
• Casey Weiner (new contributor!)

GraSPy 0.2

Highlights

This release is the result of over 8 months of work with over 25 pull requests by
10 contributors. Highlights include:

• Added AutoGMMCluster in cluster submodule. AutoGMMCluster is Python equivalent to mclust in R.
• Added subgraph submodule, which detects vertices that maximally correlates to given features.
• Added match submodule. Used for matching vertices from a pair of graphs with unknown vertex correspondence.
• Added functions for simulating a pair of correlated ER and SBM graphs.

Improvements

• Diagonal augmentation is default behavior in AdjacencySpectralEmbed.
• Added functionality in to_laplace to allow for directed graphs.
• Updated docstrings.
• Updated documentation website.
• Various bug fixes.

API Changes

• Added **kwargs argument for heatmap.

GraSPy 0.1

Highlights

This release is the result of over 2 months of work with over 18 pull requests by 3 contributors. Highlights include:

Added MultipleASE, which is a new method for embedding population of graphs.
Added mug2vec within pipieline module, which learns a feature vector for population of graphs.

GraSPy 0.0.3

Highlights

This release is the result of over 2 months of work with over 16 pull requests by
4 contributors. Highlights include:

• Optimization over covariance structures when using GaussianCluster
• Standardized sorting for visualizing graphs when using heatmap or gridplot
• Graph model classes for fitting several random graph models to input datasets
• Improved customization for heatmaps and gridplots

GraSPy 0.0.2

Highlights

This release is the result of 3 months of work with over 16 pull requests by 5 contributors. Highlights include:

• Nonparametric hypothesis testing method for testing two non-vertex matched graphs.
• Plotting updates to pairplot, gridplot and heatmaps.
• Sampling degree-correlcted stochatic block models (DC-SBM).
• import_edgelist function for importing single or multiple edgelists.
• Enforcing Black formatting for the package.

GraSPy v0.0.1

Highlights

This release is the result of over two years of work with 238 commits and 35 merges by 4 contributors.
Highlights include:

• Fast implementation of dimensionailty reduction using different implementation of SVD.
• Single and multiple graph embedding methods.
• Methods for preprocessing graphs for meaningful embeddings.
• Hypothesis testing, specifically semiparametric testing of two graphs.
• Methods for clustering vertices or population of graphs
• Plotting functions for visualization of graphs and high dimensional data.