gbm.auto (version 2024.10.01)

gbm.map: Maps of predicted abundance from Boosted Regression Tree modelling

Description

Generates maps from the outputs of gbm.step then Gbm.predict.grids, handled automatically within gbm.auto but can be run alone, and generates representativeness surfaces from the output of gbm.rsb.

Usage

gbm.map(
  x,
  y,
  z,
  byx = NULL,
  byy = NULL,
  grdfun = mean,
  mapmain = "Predicted CPUE (numbers per hour): ",
  species = "Response Variable",
  heatcolours = c("white", "yellow", "orange", "red", "brown4"),
  colournumber = 8,
  shape = NULL,
  landcol = "grey80",
  mapback = "lightblue",
  legendloc = "bottomright",
  legendtitle = "CPUE",
  lejback = "white",
  zero = TRUE,
  quantile = 1,
  byxout = FALSE,
  breaks = NULL,
  byxport = NULL,
  ...
)

Value

Species abundance maps using data provided by gbm.auto, and Representativeness Surface Builder maps using data provided by gbm.rsb, to be run in a png/par/gbm.map/dev.off sequence.

Arguments

x

Vector of longitudes, from make.grid in mapplots; x. Order by this (descending) SECOND.

y

Vector of latitudes, from make.grid in mapplots; grids[,gridslat]. Order by this (descending) first.

z

Vector of abundances generated by gbm.predict.grids, from make.grid in mapplots; grids[,predabund].

byx

Longitudinal width of grid cell, from make.grid in mapplots. Autogenerated if left blank.

byy

Latitudinal height of grid cell, from make.grid in mapplots. Autogenerated if left blank.

grdfun

make.grid operand for >=2 values per cell. Default:mean, other options: sum prod min max sd se var.

mapmain

Plot title, has species value appended. Default "Predicted CPUE (numbers per hour): ".

species

Response variable name, from basemap in mapplots; names(samples[i]). Defaults to "Response Variable".

heatcolours

Vector for abundance colour scale, defaults to the heatcol from legend.grid and draw.grid in mapplots which is c("white", "yellow", "orange" , "red", "brown4").

colournumber

Number of colours to spread heatcol over, default:8.

shape

Basemap shape to draw, from draw.shape in mapplots. Defaults to NULL which calls gbm.basemap to generate it for you. First read in a shp file e.g. myshape <- sf::st_read(dsn = paste0(savename, ".shp"), layer = savename, quiet = TRUE), then use shape = myshape.

landcol

Colour for 'null' area of map (for marine plots, this is land), from draw.shape in mapplots. Default "grey80" (light grey).

mapback

Basemap background colour, defaults to lightblue (ocean for marine plots).

legendloc

Location on map of legend box, from legend.grid in mapplots, default bottomright.

legendtitle

The metric of abundance, e.g. CPUE for fisheries, from legend.grid in mapplots. Default "CPUE".

lejback

Background colour of legend, from legend.grid in mapplots. Default "white".

zero

Force include 0-only bin in breaks.grid and thus legend? Default TRUE.

quantile

Set max quantile of data to include in bins, from breaks.grid in mapplots; lower to e.g. 0.975 cutoff outliers; default 1.

byxout

Export byx to use elsewhere? Default:FALSE.

breaks

Vector of breakpoints for colour scales; default blank, generated automatically.

byxport

Dummy param for package testing for CRAN, ignore.

...

Additional arguments for legend.grid's ... which passes to legend.

Author

Simon Dedman,

Hans Gerritsen

Details

Superseded by gbm.mapsf on 2023-08-07, but still works.

Errors and their origins:

Error in seq.default(xlim[1], xlim[2], by = byx):wrong sign in 'by' argument Check that your lat & long columns are the right way around. Ensure grids data are gridded, i.e. they are in a regular pattern of same/similar lines of lat/lon, even if they're missing sections.

Suggested parameter values: z = rsbdf[,"Unrepresentativeness"]

mapmain = "Unrepresentativeness: "

legendtitle = "UnRep 0-1"

Examples

# \donttest{
# Not run: downloads and saves external data.
# Suggested code for outputting to png:
data(grids)
# set working directory somewhere suitable
png(filename = "gbmmap.png", width = 7680, height = 7680, units = "px",
pointsize = 192, bg = "white", res = NA, family = "", type = "cairo-png")
par(mar = c(3.2,3,1.3,0), las = 1, mgp = c(2.1,0.5,0), xpd = FALSE)
gbm.map(x = grids[,"Longitude"], y = grids[,"Latitude"], z = grids[,"Effort"]
, species = "Effort")
dev.off()
# }