Introduction

ghypernet is an OpenSource R package that allows to estimate and work with Generalised Hypergeometric Ensembles of Random Graphs (gHypEG).

ghypernet has been developed specifically for the analysis of networks characterised by a large number of repeated edges. It provides efficient methods to perform hypothesis testing and model selection on such data.

Explore the provided Vignettes for some examples on how to analyse networks with ghypernet.

Installation

# Install ghypernet from CRAN
install.packages("ghypernet")

# Or the development version from GitHub:
devtools::install_github("gi0na/r-ghypernet")

Dependencies

The package uses the library BiasedUrn to work with Wallenius' non-central hypergeometric distribution. Although this is not required, it is recommended to install the BiasedUrn R package, increasing the number of 'colours', i.e., the number of dimensions of the distribution. It can be easily done modifying the makevar file. In case the BiasedUrn library cannot be found, all computations will be performed using the multinomial approximation.

References

The theoretical foundation of the generalised hypergeometric ensemble, gHypEGs, has been developed in the following works:

Casiraghi, G., Nanumyan, V., Scholtes, I., & Schweitzer, F. (2016). Generalized Hypergeometric Ensembles: Statistical Hypothesis Testing in Complex Networks. ArXiv Preprint ArXiv:1607.02441.

Casiraghi, G. (2017). Multiplex Network Regression: How do relations drive interactions?. ArXiv Preprint ArXiv:1702.02048, 15.

Casiraghi, G., Nanumyan, V., Scholtes, I., & Schweitzer, F. (2017). From Relational Data to Graphs: Inferring Significant Links Using Generalized Hypergeometric Ensembles (Vol. 10540, pp. 111–120). Springer Verlag.

Casiraghi, G. (2019). The block-constrained configuration model. Applied Network Science, 4(1), 123.

Brandenberger, L., Casiraghi, G., Nanumyan, V., & Schweitzer, F. (2019). Quantifying triadic closure in multi-edge social networks. Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, 307–310.

Casiraghi, G., & Nanumyan, V. (2021). Configuration models as an urn problem. Sci Rep 11, 13416.

Casiraghi, G. (2021) The likelihood-ratio test for multi-edge network models. J. Phys. Complex. 2 035012.

Acknowledgements

The research and development behind ghypernet is performed at the Chair of Systems Design, ETH Zürich.

Contributors

Giona Casiraghi (project lead)

Vahan Nanumyan

Laurence Brandenberger

Copyright

ghypernet is licensed under the GNU Affero General Public License.

Resolving block IDs in bccm

When printing bccm fits, the entries of the block-Omega matrix are printed in a list. There, pairs of blocks are enumerated as a cryptic number that is difficult to interpret in terms of the block assignment vector that the user has passed when fitting the bccm. The function internally represents each block by a prime number and creates the enumeration of the block-pairs as the product of the two primes of the involved blocks. However, the user usually doesn't know which prime number is assigned to which block.

For example:

library(ghypernet)

data("adj_karate")
data("vertexlabels")

bccm(adj = adj_karate, labels = vertexlabels, directed = FALSE, selfloops = FALSE)

which prints

...
Coefficients:
    Estimate    Std.Err t value    Pr(>t)    
1 1.00000000 0.00000000     Inf < 2.2e-16 ***
2 0.09044494 0.00021692  416.95 < 2.2e-16 ***
4 0.91049902 0.00040434 2251.79 < 2.2e-16 ***
...

Can the output state the block labels instead of the prime products (1, 2, 4) to identify the pairs? For example, something like

1:1
1:2
2:2

where 1 and 2 are the block-labels from vertexlabels.

gi0na / r-ghypernet Goto Github PK

r-ghypernet's Introduction

Introduction

Installation

Dependencies

References

Acknowledgements

Contributors

Copyright

r-ghypernet's People

Contributors

Stargazers

Watchers

Forkers

r-ghypernet's Issues

Recommend Projects

Recommend Topics

Recommend Org