cpPermuteEntropy {CliquePercolation} | R Documentation |

Function for determining confidence intervals of entropy values calculated for community partition from clique percolation based on randomly permuted networks of original network.

cpPermuteEntropy( W, cpThreshold.object, n = 100, interval = 0.95, CFinder = FALSE, ncores, seed = NULL )

`W` |
A qgraph object or a symmetric matrix; see also qgraph |

`cpThreshold.object` |
A cpThreshold object; see also cpThreshold |

`n` |
number of permutations (default is 100) |

`interval` |
requested confidence interval (larger than zero and smaller 1; default is 0.95) |

`CFinder` |
logical indicating whether clique percolation for weighted networks should be performed as in CFinder ; see also cpAlgorithm |

`ncores` |
Numeric.
Number of cores to use in computing results.
Defaults to |

`seed` |
Numeric.
Set seed for reproducible results.
Defaults to |

The function generates `n`

random permutations of the network
specified in `W`

. For each randomly permuted network, it runs `cpThreshold`

(see cpThreshold for more information) with `k`

and `I`

values
extracted from the cpThreshold object specified in `cpThreshold.object`

.
Across permutations, the confidence intervals of the entropy values are determined
for each `k`

separately.

The confidence interval of the entropy values is determined separately for each `k`

.
This is because larger `k`

have to produce less communities on average,
which will decrease entropy. Comparing confidence intervals of smaller `k`

to
those of larger `k`

would therefore be disadvantageous for larger `k`

.

In the output, one can check the confidence intervals of each `k`

. Moreover,
a data frame is produced that takes the cpThreshold object that was specified in
`cpThreshold.object`

and removes all rows that do not exceed the upper bound of the
confidence interval of the respective `k`

.

A list object with the following elements:

- Confidence.Interval
a data frame with lower and upper bound of confidence interval for each

`k`

- Extracted.Rows
rows extracted from

`cpThreshold.object`

that are larger than the upper bound of the specified confidence interval for each`k`

- Settings
user-specified settings

Jens Lange, lange.jens@outlook.com

## Example with fictitious data # create qgraph object W <- matrix(c(0,1,1,1,0,0,0,0, 0,0,1,1,0,0,0,0, 0,0,0,0,0,0,0,0, 0,0,0,0,1,1,1,0, 0,0,0,0,0,1,1,0, 0,0,0,0,0,0,1,0, 0,0,0,0,0,0,0,1, 0,0,0,0,0,0,0,0), nrow = 8, ncol = 8, byrow = TRUE) W <- Matrix::forceSymmetric(W) W <- qgraph::qgraph(W) # create cpThreshold object cpThreshold.object <- cpThreshold(W = W, method = "unweighted", k.range = c(3,4), threshold = "entropy") # run cpPermuteEntropy with 100 permutations and 95% confidence interval results <- cpPermuteEntropy(W = W, cpThreshold.object = cpThreshold.object, n = 100, interval = 0.95, ncores = 1, seed = 4186) # check results results ## Example with Obama data set (see ?Obama) # get data data(Obama) # estimate network net <- qgraph::EBICglasso(qgraph::cor_auto(Obama), n = nrow(Obama)) # create cpThreshold object threshold <- cpThreshold(net, method = "weighted", k.range = 3:4, I.range = seq(0.1, 0.5, 0.01), threshold = "entropy") # run cpPermuteEntropy with 50 permutations and 99% confidence interval permute <- cpPermuteEntropy(net, cpThreshold.object = threshold, interval = 0.99, n = 50, ncores = 1, seed = 4186) # check results permute

[Package *CliquePercolation* version 0.3.0 Index]