Skip to contents

When each row already describes a point, pass the table straight to hchart(). Use a data_to_* helper when you need to calculate boxplot statistics, count transitions or reorganize rows into the structure a chart expects.

What you have Helper What to do with the result
Several measurements per observation data_to_parallel_coordinates() Add one line per row with hc_add_series_list().
A numeric measurement and optional groups data_to_boxplot() Add the boxes and optional outlier series with hc_add_series_list().
One column per stage in a journey data_to_sankey() Pass the weighted links to hchart(links, type = "sankey").
Nested groups and a numeric size data_to_hierarchical() Pass the nodes to hchart(nodes, type = "treemap") or hchart(nodes, type = "sunburst").
Coordinates and text for labels data_to_annotation_labels() Pass the labels to hc_add_annotation() on an existing chart.
library(highcharter)
hc_set_theme("hcrt", colors = hc_theme("smpl")$colors)

Parallel coordinates

Each flower in iris becomes a line crossing four measurement axes. Select the numeric columns in display order. Use translucent lines to reveal where flowers overlap. The row number identifies each specimen, and color distinguishes the three species.

measurements <- iris[1:4]
series <- data_to_parallel_coordinates(
  measurements,
  name = paste(iris$Species, seq_len(nrow(iris)), sep = " · "),
  color = hex_to_rgba(colorize(iris$Species), alpha = 0.36)
)

highchart() |>
  hc_chart(type = "line", parallelCoordinates = TRUE) |>
  hc_xAxis(categories = names(measurements)) |>
  hc_add_series_list(series) |>
  hc_tooltip(
    formatter = JS(
      "function () {
    const axes = this.series.xAxis.categories;
    const rows = this.series.points.map(p => {
      const value = p.isNull ? '—' : Highcharts.numberFormat(p.y, 1) + ' cm';
      const row = axes[p.x] + ': ' + value;
      return p.x === this.x ? '<b>' + row + '</b>' : row;
    });
    return ['<b>' + this.series.name + '</b>', ...rows].join('<br>');
  }"
    )
  )

Hover over any axis to see the four measurements for that flower. The tooltip reads the points already stored in its series and emphasizes the current axis. Highcharts scales each axis independently. A constant column keeps its value, and a missing measurement leaves a gap in that flower’s line.

Boxplots

To compare petal lengths across species, let the helper calculate one box per group. With add_outliers = TRUE, it also returns a scatter series for the observations beyond the whiskers. Click the legend to hide or show the boxes and their outliers together.

boxes <- data_to_boxplot(
  iris,
  Petal.Length,
  Species,
  add_outliers = TRUE,
  name = "Petal length",
  color = hc_theme("smpl")$colors[[1]]
)

highchart() |>
  hc_chart(style = list(fontFamily = "IBM Plex Sans", fontWeight = "300")) |>
  hc_xAxis(categories = levels(iris$Species)) |>
  hc_yAxis(title = list(text = "Petal length (cm)")) |>
  hc_legend(itemStyle = list(fontWeight = "300")) |>
  hc_add_series_list(boxes)

Which airlines connected New York’s three airports to the Midwest? Each row is one flight with a recorded arrival in January 2013 at one of six selected destinations. Order the columns as origin → airline → destination. data_to_sankey() counts the flights between each pair of adjacent stages. For example, the links into Chicago O’Hare (ORD) sum to 1,230 arrivals.

flights <- read.csv("data/nyc-midwest-flights.csv.gz")
links <- data_to_sankey(flights[c("origin", "carrier", "dest")])

hchart(links, "sankey", name = "Flights", nodeWidth = 14, nodePadding = 12) |>
  hc_title(text = "From New York to the Midwest") |>
  hc_subtitle(text = "3,445 flights · January 2013") |>
  hc_tooltip(valueDecimals = 0)

Flight data from nycflights13 · Download the three flight columns. The same layout can display parallel sets in D3.

Hierarchy nodes

Count Pokémon by primary and secondary type. data_to_hierarchical() turns each primary type into a parent node and each observed combination into a leaf. Pokémon with a single type get their own leaf, such as “Only Electric”.

In this treemap from Showcase, the area of each rectangle shows the number of Pokémon. Click a primary type to explore its secondary types.

library(highcharter)
library(dplyr)

counts <- pokemon |>
  count(type_1, type_2, type_1_color, name = "n") |>
  mutate(type_2 = coalesce(type_2, paste("Only", type_1)))

nodes <- data_to_hierarchical(counts, c(type_1, type_2), n)
palette <- setNames(pokemon$type_1_color, pokemon$type_1)
nodes <- lapply(nodes, function(node) {
  if (node$level == 1L) {
    node$color <- unname(palette[node$name])
  }
  node$name <- tools::toTitleCase(node$name)
  node
})

label_style <- list(fontFamily = "Shadows Into Light", fontSize = "15px")

hchart(
  nodes,
  type = "treemap",
  name = "All Pokémon",
  allowTraversingTree = TRUE,
  breadcrumbs = list(
    buttonTheme = list(style = label_style),
    separator = list(style = label_style)
  ),
  dataLabels = list(
    style = c(
      label_style,
      list(fontWeight = "normal", color = "contrast", textOutline = "none")
    )
  ),
  layoutAlgorithm = "squarified",
  levelIsConstant = FALSE,
  levels = list(list(level = 1, dataLabels = list(enabled = TRUE)))
) |>
  hc_title(
    text = "Pokémon by type",
    style = list(fontFamily = "Shadows Into Light", fontWeight = "normal")
  ) |>
  hc_tooltip(style = label_style) |>
  hc_size(height = 480)

Annotation labels

Stage 8 of the 2017 Tour de France has three kinds of landmarks: towns, summits and climbs. Each row supplies its distance (x), elevation (y) and label (text). Convert each table with the same helper, then give its labels a placement suited to the terrain. Here 12 labels share a single elevation series.

Original annotated stage profile · Download the elevation profile (CSV).

stage <- read.csv("data/tour-2017-stage-8.csv")
towns <- data.frame(
  x = c(27.98, 45.5, 63, 84, 129.5, 159),
  y = c(255, 611, 651, 789, 382, 443),
  text = c(
    "Arbois",
    "Montrond",
    "Mont-sur-Monnet",
    "Bonlieu",
    "Chassal",
    "Saint-Claude"
  )
)
climbs <- data.frame(
  x = c(96.2, 134.5, 172.2),
  y = c(783, 540, 925),
  text = c("6.1 km · 4.6%", "7.6 km · 5.2%", "11.7 km · 6.4%")
)
summits <- data.frame(
  x = c(101.44, 138.5, 176.4),
  y = c(1026, 748, 1202),
  text = c(
    "Col de la Joux",
    "Côte de Viry",
    "Montée de la Combe<br>de Laisia Les Molunes"
  )
)

hchart(
  stage,
  "area",
  hcaes(distance, elevation),
  name = "Elevation",
  fillOpacity = 0.25
) |>
  hc_title(text = "Tour de France 2017 · Stage 8") |>
  hc_xAxis(title = list(text = "Distance (km)")) |>
  hc_yAxis(title = list(text = "Elevation (m)")) |>
  hc_tooltip(valueSuffix = " m") |>
  hc_add_annotation(
    labels = data_to_annotation_labels(towns),
    labelOptions = list(
      verticalAlign = "top",
      y = 10,
      backgroundColor = "rgba(255,255,255,0.85)"
    )
  ) |>
  hc_add_annotation(labels = data_to_annotation_labels(summits)) |>
  hc_add_annotation(
    labels = data_to_annotation_labels(climbs),
    labelOptions = list(shape = "connector", align = "right", x = -8, y = -32)
  )