PomaCorr {POMA} | R Documentation |
This function returns different correlation plots (correlogram and network plots) and a table with all pairwise correlations in the data.
PomaCorr( data, method = "pearson", shape = "square", type = "full", show_corr = FALSE, low = "#336B87", outline = "white", high = "#EA8620", label_size = 12, corr_type = "cor", coeff = 0.7 )
data |
A MSnSet object. First |
method |
Character indicating which correlation coefficient has to be computed. Options are "pearson" (default), "kendall" and "spearman". |
shape |
Character ingicating shape of correlogram. Options are "square" (default) and "circle". |
type |
Character indicating type of correlogram. Options are "full" (default), "lower" or "upper". |
show_corr |
Logical indicating if correlation coefficient for each pair of features should be plotted in correlogram or not (default = FALSE). Only recomended for a low number of features. |
low |
Colour for low end of the gradient in correlogram. |
outline |
Colour for the outline of the gradient in correlogram. |
high |
Colour for high end of the gradient in correlogram. |
label_size |
Numeric indicating label size in correlogram. |
corr_type |
Type of network to be made with correlation matrix. Options are "cor" (for global correlations) and "glasso" (for gaussian graphical model). Default is "cor". See |
coeff |
Numeric indicatin correlation coefficient. Edges with absolute weight below this value will be removed from the network. If "corr_type" is set to "glasso", this parameter indicates the regularization parameter for lasso (rho = 0 means no regularization). See |
A list with the results.
Pol Castellano-Escuder
Jerome Friedman, Trevor Hastie and Rob Tibshirani (2019). glasso: Graphical Lasso: Estimation of Gaussian Graphical Models. R package version 1.11. https://CRAN.R-project.org/package=glasso
data("st000284") # pearson correlation PomaCorr(st000284)$correlations PomaCorr(st000284)$corrplot # gaussian graphical model # library(ggraph) # PomaCorr(st000284, corr_type = "glasso")