R/dbpf_kml.R

In geocryology/PermafrostDB: PermafrostDB

# =============================================================================
#'
#' @title Write a kml plotting site locations
#'
#' @description Exports site locations as a kml file.
#'
#' @details Function needs to be extended to allow:
#'          --> filtering by location, etc.
#'          --> display of summary information per site
#'          --> assigning shape and colour of icon based on attributes
#' 
#' @param con Database connection object, as returned by \code{\link{dbpf_con}}
#' 
#' @param kmlfile Filename for kml to be produced.
#'
#' @param location_name (optional) Character, one or more location names to
#' convert.
#'
#' @param bounding_box (optional) numeric vector of the form
#'  (xmin, xmax, ymin, ymax) specifying longitude (x) and latitude (y) values
#'  to filter sites
#'
#' @return No value is returned.
#'
#' @export
#' @examples
#' \dontrun{
#' con <- dbpf_con()
#' dbpf_kml(con, kmlfile = "~/Desktop/permafrost.kml")
#' }
#' @author Stephan Gruber <stephan.gruber@@carleton.ca>
# =============================================================================

dbpf_kml <- function(con, kmlfile, location_name, bounding_box) {
  lon <- lat <- NULL  # to prevent error "no visible binding for global variable ..."
  
  for (package in c("sp", "rgdal", "plotrix", "plotKML", "rgdal", "sp", "dismo",
                      "pixmap", "RSAGA", "aqp", "stars", "colorRamps")){
      if (!requireNamespace(package, quietly = TRUE)){
        utils::install.packages(package)
      }
    }

    if (missing(con)){
      con <- dbpf_con()
    }

    #get data and filter (if filtering parameters provided)
    loc <- dbpf_locations(con)

    if (!missing(location_name)){ # filter by location name
      loc <- loc[loc$name %in% location_name,]
    }

    if (!missing(bounding_box)){ # filter by coordinates
      loc <- loc[loc$lon > bounding_box[1] &
                   loc$lon < bounding_box[2] &
                   loc$lat > bounding_box[3] &
                   loc$lat < bounding_box[4],]
    }

    # make data to include
    df <- loc[,-which(names(loc) %in% c('id', 'record_observations'))]

    #make spatial points
    spdf <- sp::SpatialPointsDataFrame(subset(loc, select=c("lon","lat")),
                                       subset(loc, select=c(-lon,-lat)))
    sp::proj4string(spdf) <- sp::CRS("+init=epsg:4326")

    #write kml
    #TODO: give table: https://github.com/cran/plotKML/blob/master/R/layer.SpatialPoints.R
    # also see http://gsif.isric.org/doku.php?id=wiki:tutorial_plotkml
    plotKML::kml_open(kmlfile)
    plotKML::kml_layer.SpatialPoints(spdf,
                            points_names = loc$name,
                            balloon=T,  # show metadata in balloon
                            #size=0.65,
                            #shape="http://maps.google.com/mapfiles/kml/pal2/icon18.png",
                            colour='black')
    plotKML::kml_close(kmlfile)

    # feedback
    print('===> KML does not contain plots (ploygon locations), feature needs to be added to code.')
}