Vector data I/O

• The library {sf} can import data from various sources:
  1. A multitude of file formats from disk (powered gdal / ogr2ogr)
  2. Various geospatial databases such as PostGIS
  3. From other sources (such as WKT, WKB other geospatial R-libraires)
• We will only introduce the first method, since this is the most common use case

GDAL / ogr2ogr

GDAL:

• is an open source translator library for raster and vector geospatial data formats.
• stands for Geospatial Data Abstraction Library
• is used in most geospatial software, be it FOSS or proprietary. The list includes: ArcGIS, QGIS, R (sf) and Python (geopandas)

GDAL Vector drivers (read)

• Since GDAL supports a long list of different geospatial file formats, most (all?) are in turn supported by {sf}
• Run the function st_drivers for a full list
• {sf} tries to guess the correct driver based on the file extension (see below)

# sf uses the geojson driver, based on the file extension
# (data from moodle →  Spatial Databases II → Database Files)
stationen_schweiz <- read_sf("data/uebung1/stationen_schweiz.geojson")

# sf uses the shapefile driver, based on the file extension
# (data from moodle →  Spatial Analysis I → Exercise Data) 
hoheitsgebiet <- read_sf("data/Spatial_Analysis_I/Hoheitsgebiet/HOHEITSGEBIET_FR.shp")

• Note: The read_sf() function is a wrapper around the st_read() function, which is the actual function that reads the data. I use read_sf since this wrapper’s default value for the quiet argument is TRUE (less verbose)

Multiple layers in one file

• Some file formats, e.g. Geopackages (*.gpkg) or Geodatabases (*.gdb) support multiple datasets in a single file.
• If no specific layer is requested read_sf() will import the first available layer
• If more than 1 layers are available, read_sf() will return a warning
• The function st_layers() will list all available layers

tlm3d_path <- "data/Spatial_Analysis_II/swiss_TLM3D.gpkg"

# Note the warning:
tlm3d <- read_sf(tlm3d_path)

Warning in CPL_read_ogr(dsn, layer, query, as.character(options), quiet, :
automatically selected the first layer in a data source containing more than
one.

---

# This will list all layers, including some metadata
st_layers(tlm3d_path)

Driver: GPKG 
Available layers:
    layer_name        geometry_type features fields       crs_name
1       tlm_bb     3D Multi Polygon    49321     13 CH1903+ / LV95
2 tlm_strassen 3D Multi Line String   191644     27 CH1903+ / LV95

tlm_bb <- read_sf(tlm3d_path, "tlm_bb")

SQL queries during import

• read_sf() understands an Spatialite SQL query provided in the query = argument
• In case query = is used, the layers = argument should be skipped (since the layers is specified in the query)

tlm_seen <- read_sf(
  tlm3d_path, 
  query = "SELECT objektart, geom FROM tlm_bb WHERE objektart = 'Stehende Gewaesser'"
  )

GDAL Vector drivers (write)

• The function st_write() is used to export an sf object to file
• Most vector drivers support reading and writing (see st_drivers / column write)
• Many file formats support appending to the dataset (see append =)

st_write(tlm_seen, "data-out/seen.geojson")

Writing layer `seen' to data source `data-out/seen.geojson' using driver `GeoJSON'
Writing 1189 features with 1 fields and geometry type 3D Multi Polygon.

🪕 Tasks

1. Import the datasets from Spatial Analysis I and II using R
2. Explore these datasets by printing the sf objects to the console
3. Try different visualization methods (see Plotting simple features)