plot_ktable: Plotting Functions

In ilangurudev/approxmapR: ApproxMAP (APPROXimate Multiple Alignment Pattern)

Plotting Functions

Description

A function which allows one to plot a cluster validation measure against the corresponding K value from the table that is produced from the find_optimal_k function.

Usage

  plot_ktable <- (ktable, validation_measure = 'silhouette', save_graph = TRUE, graph_file_name = NULL,
                          size_width = 855, size_height = 317, output_directory = "~")

Arguments

ktable	A dataframe that has been produced from find_optimal_k which contains various K values and their respective cluster validation measures values.
validation_measure	The clustering validation measure to be used for plotting. The following measures are cuurently supported: * Average silhouette width (validation_measure = 'silhouette') * Dunn index (validation_measure = 'dunn') * Average distance within clusters (validation_measure = 'average_within') * Average distance between clusters (validation_measure = 'average_between') * Ratio of average distance within cluster / average distance between clusters (validation_measure = 'wb_ratio') * Half the sum of the within cluster squared dissimilarities divided by the cluster size (validation_measure = 'within_cluster_ss')
save_graph	Default value is TRUE which will save the plot as a PNG file.
graph_file_name	Allows user to specify the file name for the graph that is being saved, if nothing is specified then a default file name is used.
size_width	The width of the graph.
size_height	The height of the graph.
output_directory	The path to where the exports should be placed.

Value

Returns a plot with the K value on the x-axis and the cluster validation measure on the y-axis.

Examples

  data("demo1")
  demo1 <- data.frame(do.call("rbind", strsplit(as.character(demo1$id.date.item), ",")))
  names(demo1) <- c("id", "period", "event")


  agg <- demo1 %>% aggregate_sequences(format = "%m/%d/%Y",
                                     unit = "month",
                                     n_units = 1,
                                     include_date = TRUE,
                                     summary_stats = TRUE)

  ktable <- agg %>% find_optimal_k(clustering = "k-medoids", min_k = 2, max_k = 9,
                                   use_cache = TRUE, save_table = TRUE)

  plot_ktable(ktable)

ilangurudev/approxmapR documentation built on March 22, 2022, 1:15 p.m.