Skip to contents

Interactive leaflet map for classified rasters and transitions

Source: R/dft_map_interactive.R
dft_map_interactive.Rd

Build a toggleable leaflet map from classified SpatRasters or remote COG URLs served via titiler. Optionally overlay land cover transitions from dft_rast_transition(). Includes layer control, legend, and fullscreen.

Usage

dft_map_interactive(
  x = NULL,
  aoi = NULL,
  transition = NULL,
  rgb = NULL,
  rgb_rescale = NULL,
  class_table = NULL,
  source = "io-lulc",
  titiler_url = getOption("drift.titiler_url"),
  basemaps = c(Light = "CartoDB.Positron", `Esri Satellite` = "Esri.WorldImagery",
    `Google Satellite` = "https://mt1.google.com/vt/lyrs=s&x={x}&y={y}&z={z}",
    OpenTopoMap = "OpenTopoMap"),
  legend_position = "bottomright",
  zoom = 14
)

Arguments

x

A named list of classified terra::SpatRasters (e.g. from dft_rast_classify()) or a named character vector of COG URLs. A single SpatRaster or URL string is auto-wrapped into a length-1 list/vector. Names become the layer toggle labels (years, seasons, etc.). May be NULL when rgb is supplied, for an imagery-only map.

aoi

An sf polygon for the area of interest outline. NULL (default) omits the AOI layer.

transition

Output of dft_rast_transition() — a list with elements raster (factor SpatRaster) and summary (tibble). Each transition type becomes a toggleable overlay. Stable transitions (same from/to class) are excluded by default. NULL (default) omits transition overlays.

rgb

Dated reference imagery: a named list of three-band terra::SpatRasters in red, green, blue order (the output of dft_stac_composite() with its default bands), or a named character vector of three-band COG URLs served through titiler_url. Names become the layer labels, e.g. "2017 Jun–Jul". Local rasters are drawn with leafem::addRasterRGB(), which reads every cell into memory; for a floodplain-sized composite, publish the cached COG and pass its URL instead. NULL (default) adds no imagery.

rgb_rescale

Numeric length-2 reflectance range stretched to the full display range, the same for every band of every rgb layer. When NULL, local rasters stretch each band from its 2nd to 98th percentile across all composites pooled, and COG URLs use c(0, 0.3), a range that suits vegetated land in surface reflectance.

class_table

A tibble with columns code, class_name, color (hex). When NULL, loaded via dft_class_table() using source.

source

Character. Used to load a shipped class table when class_table is NULL. One of "io-lulc" or "esa-worldcover".

titiler_url

Base URL of a titiler instance (e.g. "https://titiler.example.com"). Only used when x contains COG URLs. Defaults to getOption("drift.titiler_url"). If NULL in COG mode, an error is raised prompting the user to set the option.

basemaps

Named character vector of provider tile IDs or tile URL templates (starting with http). The first element is the default basemap. Names become radio button labels.

legend_position

Legend placement passed to leaflet::addLegend(). Set to NULL to suppress the legend.

zoom

Initial zoom level.

Value

A leaflet::leaflet htmlwidget. The first layer in x is visible by default; other layers are hidden but toggleable. Transition overlays are visible by default when supplied. rgb layers are hidden when x is supplied (toggle them on beneath it); without x the first is shown.

Details

When only x is supplied, classified layers appear as radio-toggle overlays (one visible at a time). When transition is also supplied, each transition type (e.g. Trees -> Rangeland) is added as a checkbox overlay that can be shown simultaneously on top of any classified layer.

When rgb is supplied, each composite (e.g. from dft_stac_composite()) is added as a switchable overlay beneath the classified and transition layers, so dated imagery can be toggled under a land-cover label to check it. Every composite gets the same stretch: each band's display range is set once, from all the composites pooled, so a brightness or colour difference between two years is in the data and not an artefact of stretching each image to its own histogram.

Examples

# Single classified raster — returns a leaflet widget
r <- terra::rast(system.file("extdata", "example_2020.tif", package = "drift"))
classified <- dft_rast_classify(r, source = "io-lulc")
map <- dft_map_interactive(classified)
class(map)
#> [1] "leaflet"    "htmlwidget"

# Multiple years with AOI — toggle between time periods
aoi <- sf::st_read(
  system.file("extdata", "example_aoi.gpkg", package = "drift"),
  quiet = TRUE
)
files <- c("2017" = "example_2017.tif", "2020" = "example_2020.tif",
           "2023" = "example_2023.tif")
rasters <- lapply(files, function(f) {
  terra::rast(system.file("extdata", f, package = "drift"))
})
classified <- dft_rast_classify(rasters, source = "io-lulc")
map <- dft_map_interactive(classified, aoi = aoi)
if (interactive()) map

# Combined: classified layers + transition overlays
trans <- dft_rast_transition(classified, from = "2017", to = "2023",
                             from_class = "Trees")
map <- dft_map_interactive(classified, aoi = aoi, transition = trans)
if (interactive()) map

if (FALSE) { # \dontrun{
# Remote COGs via titiler (requires options(drift.titiler_url = "..."))
cogs <- c("2017" = "https://bucket.s3.amazonaws.com/lulc_2017.tif",
          "2023" = "https://bucket.s3.amazonaws.com/lulc_2023.tif")
dft_map_interactive(cogs, source = "io-lulc")

# Dated true-colour composites beneath the land cover (network + gdalcubes)
tc <- dft_stac_composite(aoi, years = c(2017, 2023), months = 6:7)
dft_map_interactive(classified, aoi = aoi, rgb = tc)
} # }