15  Libraries

By creating an R package, we document functions and make them easily available across workflows and apps.

• msens
  R library of functions for mapping marine sensitivities, sponsored by BOEM

Functions read data from Chapter 12, analyze it, and visualize results. The Shiny apps in Chapter 16 consume msens::* helpers exclusively — app code never talks to the tile factory or PMTiles URL templates directly.

15.1 Map-tile helpers (as of 0.3.4)

The following helpers bridge the Shiny app to the msens TiTiler factory (Chapter 14):

Helper	Purpose
cell_tile_url(sql, colormap, rescale, color, mtime, base)	Compose an XYZ tile URL template for the factory, base64url-encoding sql and omitting colormap/rescale when color= is set (single-color mask).
cell_stats(sql, mtime, base)	GET /msens/statistics — returns list(n, min, max, mean, std, p2, p50, p98) for a stable legend rescale.
add_cell_tiles(m, tile_url, ...)	Wraps mapgl::add_raster_source() + mapgl::add_raster_layer() so the Shiny app adds a cell-values (or mask) layer to a map in one call.
add_cells(m, r, colors, ...) (retained)	Legacy image-source path for static report figures that need terra SpatRaster rendering (e.g., Quarto-to-DOCX output that can’t reach the tile server).

The SQL passed to cell_tile_url() and cell_stats() must return exactly two columns, cell_id (integer) and value (numeric) — see the examples in Chapter 12.