Rainbow of Gower's similarity values for each study (Transitivity evaluation)

Illustrating the range of Gower's similarity values for each study in the network.

Usage

rainbow_similarities(
  results,
  axis_title_size = 12,
  axis_text_size = 12,
  label_size = 3.5
)

Arguments

results: An object of S3 class comp_clustering. See 'Value' in comp_clustering.
axis_title_size: A positive integer for the font size of axis title (both axes). axis_title_size determines the axis.title argument found in the theme's properties in the R-package ggplot2.
axis_text_size: A positive integer for the font size of axis text (both axes). axis_text_size determines the axis.text argument found in the theme's properties in the R-package ggplot2.
label_size: A positive integer for the font size of labels appearing on each study-specific segment. label_size determines the size argument found in the geom's aesthetic properties in the R-package ggplot2.

Value

A horizontal bar plot illustrating the range of Gower's similarity values for each study using shades of red and green to indicate low and substantial between-comparisons similarity, respectively: the darker the red, the lower the between-comparisons similarity (corresponding to values close to 0), whilst the darker the green, the higher the between-comparisons similarity (corresponding to values close to 1). The study names appear on the y-axis in the order they appear in results and the similarity values appear on the x-axis. Red and blue points refer to the (average) within-comparison and between-comparisons similarity, respectively, for each study.

Details

The range of Gower's similarity values for each study result from calculating the Gower's dissimilarity of a study versus the remaining studies in the network for a set of clinical and methodological characteristics that may act as effect modifiers. Then, the Gower's dissimilarities are transformed into similarities by subtracting each value from 1: Gower's dissimilarities take values from 0 to 1, with 0 and 1 implying perfect similarity and perfect dissimilarity, respectively.

The exact similarity values appear as dotted, vertical, black lines on each bar. The root mean square of the similarity values (fixed weights) also appear on each bar as a black diamond.

References

Gower J. General Coefficient of Similarity and Some of Its Properties. Biometrics 1971;27(4):857–71. doi: 10.2307/2528823

Author

Loukia M. Spineli