Map layer that draws rasters. Supported visual variable is: col (the color).
tm_raster(
col = tm_vars(),
col.scale = tm_scale(),
col.legend = tm_legend(),
col.chart = tm_chart_none(),
col.free = NA,
col_alpha = tm_const(),
col_alpha.scale = tm_scale(),
col_alpha.legend = tm_legend(),
col_alpha.chart = tm_chart_none(),
col_alpha.free = NA,
zindex = NA,
group = NA,
group.control = "check",
blend = "over",
options = opt_tm_raster(),
...
)opt_tm_raster(interpolate = FALSE)
Visual variable that determines the color. See details. Unit: Color -- a color name, hex string.
Visual variable that determines the color transparency. See details. Unit: Proportion -- numeric 0-1 (0 = fully transparent, 1 = fully opaque).
Controls the stacking order of map layers. Should be set to a value above 400. By default, layers are stacked in call order, starting at 401. See details.
Name of the group to which this layer belongs. This is only
relevant in view mode, where layer groups can be switched (see group.control)
In view mode, the group control determines how layer groups
can be switched on and off. Options: "radio" for radio buttons
(meaning only one group can be shown), "check" for check boxes
(so multiple groups can be shown), and "none" for no control
(the group cannot be (de)selected).
Compositing operator for layer blending. Default "over" applies
no blending. See the "Layer blending" section for the supported values.
options passed on to the corresponding opt_<layer_function> function
to catch deprecated arguments from version < 4.0
Should the raster image be interpolated? Currently only applicable in view mode (passed on to grid)
The visual variable arguments (e.g. col) can be specified with a data
variable name (e.g., a spatial vector attribute or a raster layer of the object
specified in tm_shape()), with a visual value (for col, a color is expected), or with a geometry-derived variable (see below).
See vignette: Visual variables.
Multiple values can be specified: in that case facets are created.
These facets can be combined with other faceting data variables, specified with tm_facets().
See vignette: Facets.
The *.scale arguments determine the used scale to map the data values to
visual variable values. These can be specified with one of the available
tm_scale_*() functions. The default is specified by the tmap option (tm_options()) scales.var.
See `rvignette: Scales
The *.legend arguments determine the used legend, specified with tm_legend().
The default legend and its settings are determined by the tmap options (tm_options()) legend. .
See `rvignette: Legends
The *.chart arguments specify additional charts, specified with tm_chart_, e.g. tm_chart_histogram().
See `rvignette: Charts
The *.free arguments determine whether scales are applied freely across facets, or shared.
A logical value is required. They can also be specified with a vector of three
logical values; these determine whether scales are applied freely per facet dimension.
This is only useful when facets are applied (see tm_facets()).
There are maximally three facet dimensions: rows, columns, and pages. This only
applies for a facet grid (tm_facets_grid()). For instance, col.free = c(TRUE, FALSE, FALSE)
means that for the visual variable col, each row of facets will have its own
scale, and therefore its own legend. For facet wraps and stacks
(tm_facets_wrap() and tm_facets_stack()) there is only one facet dimension,
so the *.free argument requires only one logical value.
Currently, three geometry-derived variables are implemented:
"AREA" (polygons only), which uses the feature area;
"LENGTH" (lines only), which uses the feature length; and
"MAP_COLORS", which assigns values so that adjacent features receive
different values, making it particularly suitable for coloring
neighbouring polygons.
Note that geometry-derived variables do not generate a legend automatically.
If a legend is required, compute the corresponding variable explicitly,
for example with sf::st_area(), sf::st_length(), or
tmaptools::map_coloring(), and use the resulting values instead.
Every visual variable maps data values to a specific output unit.
Knowing the unit matters when supplying constant values via tm_const(),
or output ranges via values.range / values.scale in the scale
functions.
| Variable | Output unit | Notes |
fill, col, bgcol | color | name, hex, or palette string |
fill_alpha, col_alpha, bgcol_alpha | proportion 0-1 | 0 = transparent, 1 = opaque |
size (symbols, bubbles, squares, dots) | typographic lines | 1 line approx. 1/6 inch; scaled by values.scale |
size (circles) | meters | plain numeric or a units object |
size (text, labels) | multiplier | 1 = 12 pt (plot) / 12 px (view) |
lwd | lwd | base R units; 1 lwd approx. 0.75 pt at 96 dpi |
lty | -- | integer 1-6 or name ("solid", "dashed", ...) |
shape | -- | integer pch 1-25 or single character |
angle | degrees | 0-360, clockwise from north |
fontface | -- | "plain", "bold", "italic", "bold.italic" |
size in tm_symbols, tm_bubbles, tm_squares, tm_dots)
"Lines" is a typographic unit: one line is approximately 1/6 inch (the
default base line-height in R graphics). The global multiplier
tmap_options(values.scale = list(size.bubbles = 1.5)) scales all symbol
sizes without changing the data mapping.
size in tm_circles)
The value is a geographic radius in meters. A plain numeric vector is
interpreted as meters; a units object (from the units package) is
automatically converted, so units::as_units(1, "mi") gives a 1-mile
radius. Because the radius is geographic, circles scale with zoom in
interactive (view) mode -- unlike bubble symbols which keep a fixed screen
size.
size in tm_text, tm_labels)
The value is a multiplier of the base font size. size = 1 renders at
12 pt in plot mode (R's default par("ps")) and at 12 px in view
mode (gp$cex * 12 px, see tmapLeafletDataPlot.tm_data_text); the two
modes are consistent by design.
blend)
Blend modes control how a layer's pixels are combined with the pixels beneath it. For each pixel, let \(S\) be the source (top layer) RGB value and \(D\) be the destination (bottom layer) RGB value, both normalised to \([0, 1]\).
blend | Formula | Use case |
"over" | \(S \cdot \alpha + D \cdot (1 - \alpha)\) | Standard alpha compositing (default) |
"multiply" | \(S \times D\) | Hillshading over colour raster; both layers darken each other |
"screen" | \(1 - (1 - S)(1 - D)\) | Inverse of multiply; brightens |
"overlay" | multiply if \(D < 0.5\), screen if \(D \geq 0.5\) | Boosts contrast; preserves midtones |
"darken" | \(\min(S, D)\) | Keeps the darker of the two layers per channel |
"lighten" | \(\max(S, D)\) | Keeps the lighter of the two layers per channel |
Requires R >= 4.2 and a compatible graphics device (e.g.
png(type = "cairo"), svg()). In view mode, blending is applied via
CSS mix-blend-mode. See grid::groupGrob() for the full list of
supported operators.
In view mode, each layer is rendered in a Leaflet pane named "tmap{zindex}"
(e.g., "tmap401", "tmap402"), with base tile layers placed in the
standard "tile" pane.
if (FALSE) {
# load land data
data(land, World)
tm_shape(land) +
tm_raster("cover")
tm_shape(land) +
tm_raster("elevation", col.scale = tm_scale_continuous(values = terrain.colors(9))) +
tm_shape(World) +
tm_borders()
}
Run the code above in your browser using DataLab