Creates a threshold scale, where threshold breaks are computed from the given quantile
probs
.
Quantile scale legend ticks will be quantile values at each quantile break (including
limits), not the quantile probabilities at each break. This can be overridden in
tick_format
.
Example:
tick_format = function(x) as.character(probs)
.
Usage
scale_color_quantile(
col,
palette = scales::viridis_pal(),
na_color = "#000000",
probs = seq.int(0, 1, 0.25),
data = NULL,
tick_format = format_number,
col_label = "{.col}",
legend = TRUE
)
scale_quantile(
col,
range = 0:1,
na_value = 0,
probs = seq.int(0, 1, 0.25),
data = NULL,
col_label = "{.col}",
legend = TRUE
)
Arguments
- col
<
name
|string
> The name of the column containing data to be scaled. Must be a valid input torlang::ensym()
; either a named column (non-standard evaluation), a string. Supports tidy-eval.- palette
<
color
|function
> The colour palette of the colour scale. Must be a:vector of RGBA hex colours,
a palette generator function, taking a length parameter, or
a palette ramp created from
scales::colour_ramp()
A
scales::colour_ramp()
interpolator is created from the input palette.- na_color
<
color
> The colour value forNA
input values.- probs
<
numeric
> The quantile probabilities. Must be between 0 and 1.- data
<
numeric
> The data used to compute the quantiles. IfNULL
, will be taken from the layer data.- tick_format
<
function
> A label function taking a vector of ticks returning formatted ticks.- col_label
<
string
|function
> A template string or a label function for customising the column name in the legend.if
col_label
is a string,{.col}
may be used to represent thecol
nameif
col_label
is a function, the function must take a single argument: thecol
name
- legend
<
boolean
> Indicate whether the legend should be displayed for this scale.- range
<
numeric
> The output range of the numeric scale. Must be length >= 2. Astats::approxfun()
interpolator is created from the input range.- na_value
<
number
> The output value forNA
input values.
Note
As the quantiles are computed from input data, quantile scales are incompatible with
layers that load data from a url (e.g mvt_layer
). If quantiles for remote data are
known, a quantile scale can be constructed manually with scale_threshold()
.
See also
Other scales:
rescale_center()
,
rescale_diverge()
,
scale_category()
,
scale_identity()
,
scale_linear()
,
scale_log()
,
scale_power()
,
scale_quantize()
,
scale_symlog()
,
scale_threshold()