working_imagery.Rmd

---
title: "Working with imagery"
author: <a href="https://orcid.org/0000-0001-8457-4658">Diego Hernangómez <img alt="ORCID logo" src="https://info.orcid.org/wp-content/uploads/2019/11/orcid_24x24.png" width="16" height="16" /></a>
description: >
  Learn how to include images on your maps.
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  warning = FALSE,
  message = FALSE,
  dpi = 300,
  tidy = "styler",
  dev = "ragg_png",
  out.width = "100%"
)
```

**mapSpain** provides a powerful interface for working with imagery.
**mapSpain** can download static files as `.png` or `.jpeg` files (depending on
the Web Map Service) and use them along your shapefiles.

**mapSpain** also includes a plugin for **R**
[leaflet](https://rstudio.github.io/leaflet/) package, that allows you to
include several basemaps on your interactive maps.

The services are implemented via the leaflet plugin
[leaflet-providersESP](dieghernan.github.io/leaflet-providersESP/). You can
check a display of each provider on the previous link.

## Static tiles

An example of how you can include several tiles to create a static map. We focus
here on layer provided by La Rioja's [Infraestructura de Datos Espaciales
(IDERioja)](https://www.iderioja.larioja.org/).

**When working with imagery, it is important to set `moveCAN = FALSE`, otherwise
the images for the Canary Islands won't be accurate.**

```{r static1}
library(mapSpain)
library(sf)
library(ggplot2)
library(tidyterra)

# Logroño

lgn_borders <- esp_get_munic(munic = "Logroño")

# Convert to Mercator (EPSG:3857) as a general advice when working with tiles
lgn_borders <- st_transform(lgn_borders, 3857)

tile_lgn <- esp_getTiles(lgn_borders, "IGNBase.Todo", bbox_expand = 0.5)

ggplot(lgn_borders) +
  geom_spatraster_rgb(data = tile_lgn) +
  geom_sf(fill = NA, linewidth = 2, color = "blue")
```

### Alpha value on tiles

Some tiles could be loaded with or without an alpha value, that controls the
transparency of the object:

```{r static2}
madrid <- esp_get_ccaa("Madrid", epsg = 3857)

# Example without transparency

basemap <- esp_getTiles(madrid, "IGNBase.Gris", zoommin = 1)
tile_opaque <- esp_getTiles(madrid, "UnidadesAdm.Limites", transparent = FALSE)

ggplot() +
  geom_spatraster_rgb(data = basemap) +
  geom_spatraster_rgb(data = tile_opaque) +
  theme_void()
```

Now let's check the same code using the `tranparent = TRUE` option:

```{r static_transp}
# Example with transparency

tile_alpha <- esp_getTiles(madrid, "UnidadesAdm.Limites", transparent = TRUE)

# Same code than above for plotting
ggplot() +
  geom_spatraster_rgb(data = basemap) +
  geom_spatraster_rgb(data = tile_alpha) +
  theme_void()
```

Now the two tiles overlaps with the desired alpha value.

### Masking tiles

Another nice feature is the ability of masking the tiles, so more advanced maps
can be plotted:

```{r static3}
rioja <- esp_get_prov("La Rioja", epsg = 3857)

basemap <- esp_getTiles(rioja, "IGNBase.Gris", bbox_expand = 0.1, zoommin = 1)

masked <- esp_getTiles(rioja, "IGNBase.Todo", mask = TRUE, zoommin = 1)

ggplot() +
  geom_spatraster_rgb(data = basemap, maxcell = 10e6) +
  geom_spatraster_rgb(data = masked, maxcell = 10e6)
```

## Dynamic maps with Leaflet

**mapSpain** provides a plugin to be used with the `leaflet` package. Here you
can find some quick examples:

### Earthquakes in Tenerife (last year)

```{r leaflet1}
library(leaflet)

tenerife_leaf <- esp_get_nuts(
  region = "Tenerife", epsg = 4326,
  moveCAN = FALSE
)


bbox <- as.double(round(st_bbox(tenerife_leaf) + c(-1, -1, 1, 1), 2))

# Start leaflet
m <- leaflet(tenerife_leaf,
  elementId = "tenerife-earthquakes",
  width = "100%", height = "60vh",
  options = leafletOptions(minZoom = 9, maxZoom = 18)
)

# Add layers
m <- m %>%
  addProviderEspTiles("IDErioja.Relieve") %>%
  addPolygons(color = NA, fillColor = "red", group = "Polygon") %>%
  addProviderEspTiles("Geofisica.Terremotos365dias",
    group = "Earthquakes"
  )

# Add additional options
m %>%
  addLayersControl(
    overlayGroups = c("Polygon", "Earthquakes"),
    options = layersControlOptions(collapsed = FALSE)
  ) %>%
  setMaxBounds(bbox[1], bbox[2], bbox[3], bbox[4])
```

### Population density in Spain

A map showing the population density of Spain as of 2019:

```{r leaflet2}
munic <- esp_get_munic_siane(
  year = 2019,
  epsg = 4326,
  moveCAN = FALSE,
  rawcols = TRUE
)

# Get area in km2 from siane munic
# Already on the shapefile

munic$area_km2 <- munic$st_area_sh * 10000

# Get population

pop <- mapSpain::pobmun19

# Paste
munic_pop <- merge(munic, pop[, c("cmun", "cpro", "pob19")],
  by = c("cmun", "cpro"),
  all.x = TRUE
)

munic_pop$dens <- munic_pop$pob19 / munic_pop$area_km2
munic_pop$dens_label <- prettyNum(round(munic_pop$dens, 2),
  big.mark = ".",
  decimal.mark = ","
)

# Create leaflet
bins <- c(0, 10, 25, 100, 200, 500, 1000, 5000, 10000, Inf)


pal <- colorBin("inferno", domain = munic_pop$dens, bins = bins, reverse = TRUE)

labels <- sprintf(
  "<strong>%s</strong><br/><em>%s</em><br/>%s pers. / km<sup>2</sup>",
  munic_pop$rotulo,
  munic_pop$ine.prov.name,
  munic_pop$dens_label
) %>% lapply(htmltools::HTML)


leaflet(elementId = "SpainDemo", width = "100%", height = "60vh") %>%
  setView(lng = -3.684444, lat = 40.308611, zoom = 5) %>%
  addProviderEspTiles("IGNBase.Gris") %>%
  addPolygons(
    data = munic_pop,
    fillColor = ~ pal(dens),
    fillOpacity = 0.6,
    color = "#44444",
    weight = 0.5,
    smoothFactor = .1,
    opacity = 1,
    highlightOptions = highlightOptions(
      color = "white",
      weight = 1,
      bringToFront = TRUE
    ),
    popup = labels
  ) %>%
  addLegend(
    pal = pal, values = bins, opacity = 0.7,
    title = "<small>Pop. Density km<sup>2</sup></small><br><small>(2019)</small>",
    position = "bottomright"
  )
```

## Providers available

The list `esp_tiles_providers` includes the data of the available providers you
can use on functions described above. This list includes all the parameters
needed to replicate the API request. See the static url of each provider:

```{r providers, echo=FALSE}
library(reactable)

prov_names <- names(esp_tiles_providers)

urls <- vapply(esp_tiles_providers, function(x) {
  # Create url
  allpars <- x$static
  q <- allpars$q
  rest <- modifyList(
    allpars,
    list(
      attribution = NULL,
      q = NULL
    )
  )
  rest_temp <- paste0(names(rest), "=", rest, collapse = "&")
  templurl <- paste0(q, rest_temp)

  if (grepl("rioja", q, ignore.case = TRUE)) {
    templurl <- q
  }
  templurl
}, FUN.VALUE = character(1))

df <- data.frame(
  provider = prov_names,
  static_url = urls
)

rownames(df) <- NULL
reactable(df,
  filterable = TRUE,
  searchable = TRUE,
  showPageSizeOptions = TRUE,
  striped = TRUE,
  paginationType = "jump",
  elementId = "leaflet-prov-summ"
)
```