Visualize.Scale.Quantile (Visualize v0.2.35)

Quantile scale for mapping sorted data to discrete buckets.

Unlike quantize scales which divide the domain uniformly, quantile scales divide the data into buckets with equal numbers of samples. This is useful when you want each color/category to represent the same number of data points.

Examples

# Divide data into quartiles
data = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 100]

scale = Visualize.Scale.Quantile.new()
  |> Visualize.Scale.Quantile.domain(data)
  |> Visualize.Scale.Quantile.range(["Q1", "Q2", "Q3", "Q4"])

Visualize.Scale.Quantile.apply(scale, 2)   # => "Q1"
Visualize.Scale.Quantile.apply(scale, 5)   # => "Q2"
Visualize.Scale.Quantile.apply(scale, 100) # => "Q4"

# Get the quantile thresholds
Visualize.Scale.Quantile.quantiles(scale)  # => [3.5, 6.0, 8.5]

Summary

Types

t()

Functions

apply(scale, value)

Maps a value to a discrete range value based on quantiles

bandwidth(quantile)

Returns 0: a quantile scale has no bandwidth

clamp(scale, clamp?)

Identity: a quantile scale always maps into its range

domain(scale, values)

Sets the domain from a list of sample values.

invert(quantile, value)

Returns nil: a quantile scale maps many values to one bucket

invert_extent(scale, value)

Returns the extent of domain values that map to a given range value.

new()

Creates a new quantile scale

nice(scale)

Identity

padding(scale, padding)

Identity: a quantile scale has no padding

quantiles(scale)

Returns the quantile thresholds

range(scale, values)

Sets the discrete output range

scale(scale, value)

Alias of apply/2, kept for one release (spec/03 §1.3)

ticks(scale, count)

Returns the quantile thresholds (quantiles/1); the count is ignored