entropy_overall: Calculate information content of the dataset

In NetFACS: Network Applications to Facial Communication Data

Calculate information content of the dataset

Description

Compares the observed and expected information content of the dataset.

Usage

entropy_overall(x)

Arguments

x	An object of class netfacs or simply a binary matrix of 0s and 1s, with elements in columns and events in rows.

Value

Function returns a summary tibble containing the observed entropy, expected entropy and entropy ratio (observed / expected) of the dataset. Observed entropy is calculated using Shannon's information entropy formula - ∑_{i = 1}^n p_i \log (p_i). Expected entropy is based on randomization (shuffling the observed elements while maintaining the number of elements per row) and represents the maximum entropy that a dataset with the same properties as this one can reach. Ratios closer to 0 are more ordered; ratios closer to 1 are more random.

References

Shannon, C. E. (1948). A Mathematical Theory of Communication. Bell System Technical Journal. https://doi.org/10.1002/j.1538-7305.1948.tb01338.x

Examples

### how do angry facial expressions differ from non-angry ones?
data(emotions_set)
angry.face <- netfacs(
  data = emotions_set[[1]],
  condition = emotions_set[[2]]$emotion,
  test.condition = "anger",
  ran.trials = 100,
  combination.size = 2
)

entropy_overall(angry.face)

NetFACS documentation built on Dec. 7, 2022, 1:12 a.m.