parcats: create plotly parallel categories diagram from alluvial plot

In parcats: Interactive Parallel Categories Diagrams for 'easyalluvial'

create plotly parallel categories diagram from alluvial plot

Description

creates an interactive parallel categories diagram from an 'easyalluvial' plot using the 'plotly.js' library

Usage

parcats(
  p,
  marginal_histograms = TRUE,
  data_input = NULL,
  imp = TRUE,
  width = NULL,
  height = NULL,
  elementId = NULL,
  hoveron = "color",
  hoverinfo = "count+probability",
  arrangement = "perpendicular",
  bundlecolors = TRUE,
  sortpaths = "forward",
  labelfont = list(size = 24, color = "black"),
  tickfont = NULL,
  offset_marginal_histograms = 0.7,
  offset_imp = 0.9
)

Arguments

p	alluvial plot
marginal_histograms	logical, add marginal histograms, Default: TRUE
data_input	dataframe, data used to create alluvial plot, Default: NULL
imp	dataframe, with not more then two columns one of them numeric containing importance measures and one character or factor column containing corresponding variable names as found in training data.
width	integer, htmlwidget width in pixels, Default: NULL
height	integer, htmlwidget height in pixels, Default: NULL
elementId	, htmlwidget elementid, Default: NULL
hoveron	character, one of c('category', 'color', 'dimension'), Sets the hover interaction mode for the parcats diagram.', 'If 'category', hover interaction take place per category.', 'If 'color', hover interactions take place per color per category.', 'If 'dimension', hover interactions take place across all categories per dimension., Default: 'color'
hoverinfo	character, one of c('count', 'probability', 'count+probability') set info displayed on mouse hover Default: 'count+probability'
arrangement	character, one of c('perpendicular', 'freeform', 'fixed') 'Sets the drag interaction mode for categories and dimensions.', 'If 'perpendicular', the categories can only move along a line perpendicular to the paths.', 'If ‘freeform', the categories can freely move on the plane.’, 'If ‘fixed', the categories and dimensions are stationary.’, Default: 'perpendicular'
bundlecolors	logical, 'Sort paths so that like colors are bundled together within each category.', Default: TRUE
sortpaths	character, one of c('forward', 'backward'), 'Sets the path sorting algorithm.', 'If 'forward', sort paths based on dimension categories from left to right.', Default: 'forward' 'If 'backward', sort paths based on dimensions categories from right to left.'
labelfont	list, 'Sets the font for the ‘dimension' labels.’, Default: list(size = 24, color = 'black')
tickfont	list, Sets the font for the ‘category' labels.’, Default: NULL
offset_marginal_histograms	double, height ratio reserved for parcats diagram, Default: 0.8
offset_imp	double, width ratio reserved for parcats diagram, Default: 0.9

Details

most parameters are best left at default values

Value

htmlwidget

See Also

alluvial_wide , alluvial_long , alluvial_model_response , alluvial_model_response_caret

Examples

library(easyalluvial)

# alluvial wide ---------------------------------
p = alluvial_wide(mtcars2, max_variables = 5)

parcats(p, marginal_histograms = FALSE)

parcats(p, marginal_histograms = TRUE, data_input = mtcars2)

if(check_pkg_installed("randomForest", raise_error = FALSE)) {
  # alluvial for model response --------------------
  df = mtcars2[, ! names(mtcars2) %in% 'ids' ]
  m = randomForest::randomForest( disp ~ ., df)
  imp = m$importance
  dspace = get_data_space(df, imp, degree = 3)
  pred = predict(m, newdata = dspace)
  p = alluvial_model_response(pred, dspace, imp, degree = 3)

  parcats(p, marginal_histograms = TRUE, imp = TRUE, data_input = df)
}

parcats documentation built on June 22, 2024, 11:56 a.m.