Visualize.Scale.Log (Visualize v0.2.25)

A logarithmic scale for data with exponential distributions.

Useful when data spans many orders of magnitude.

Summary

Types

t()

Functions

apply(log, value)

Applies the scale to a value

bandwidth(log)

Returns 0: a log scale has no bandwidth

clamp(scale, clamp?)

Enables or disables clamping

domain(scale, list)

Sets the domain (must be positive for log scale)

invert(log, value)

Inverts the scale

new(base \\ 10)

Creates a new log scale with optional base

nice(scale)

Extends domain to nice powers of the base

padding(scale, padding)

Identity: a log scale has no padding

range(scale, list)

Sets the range

ticks(log, count)

Generates about count tick values, as d3-scale's log ticks do.

Types

t()

@type t() :: %Visualize.Scale.Log{
  base: number(),
  clamp?: boolean(),
  domain: [number()],
  range: [number()]
}

Functions

apply(log, value)

@spec apply(t(), number()) :: float()

Applies the scale to a value

bandwidth(log)

@spec bandwidth(t()) :: 0

Returns 0: a log scale has no bandwidth

clamp(scale, clamp?)

@spec clamp(t(), boolean()) :: t()

Enables or disables clamping

domain(scale, list)

@spec domain(t(), [number()]) :: t()

Sets the domain (must be positive for log scale)

invert(log, value)

@spec invert(t(), number()) :: float()

Inverts the scale

new(base \\ 10)

@spec new(number()) :: t()

Creates a new log scale with optional base

nice(scale)

@spec nice(t()) :: t()

Extends domain to nice powers of the base

padding(scale, padding)

@spec padding(t(), number()) :: t()

Identity: a log scale has no padding

range(scale, list)

@spec range(t(), [number()]) :: t()

Sets the range

ticks(log, count)

@spec ticks(t(), integer()) :: [number()]

Generates about count tick values, as d3-scale's log ticks do.

When the domain spans fewer decades than count and the base is an integer, every k · base^i (1 ≤ k < base) inside the domain is a tick — the familiar 1, 2, 3, … 9, 10, 20, 30 … — and, when that still yields fewer than count / 2 ticks, linear ticks over the domain are used instead. Otherwise the exponents are thinned to about count nice values and mapped back through the base.