Autoplots for gt_cluster_pca objects — autoplot.gt_cluster

For gt_cluster_pca, autoplot produces a plot of a metric of choice ('BIC', 'AIC' or 'WSS') against the number of clusters (k). This plot is can be used to infer the best value of k, which corresponds to the smallest value of the metric (the minimum in an 'elbow' shaped curve). In some cases, there is not 'elbow' and the metric keeps decreasing with increasing k; in such cases, it is customary to choose the value of k at which the decrease in the metric reaches as plateau. For a programmatic way of choosing k, use gt_cluster_pca_best_k().

Usage

# S3 method for class 'gt_cluster_pca'
autoplot(object, metric = c("BIC", "AIC", "WSS"), ...)

Arguments

object: an object of class gt_dapc
metric: the metric to plot on the y axies, one of 'BIC', 'AIC', or 'WSS' (with sum of squares)
...: not currently used.

Value

a ggplot2 object

Details

autoplot produces simple plots to quickly inspect an object. They are not customisable; we recommend that you use ggplot2 to produce publication ready plots.