In afsc-gap-products/akgfmaps: Alaska Groundfish and Ecosystem Survey Area Mapping

knitr::opts_chunk$set(echo = TRUE)

Introduction

This vignette demonstrates how to create IDW maps using akgfmaps. If you haven't installed akgfmaps already, follow the installation instructions on the \code{README}. Then, load the akgfmaps package.

library(akgfmaps)

The package and all of the dependencies should load. If they do not, make sure you are running R >= 3.6, and verify that all of the dependencies listed in the \code{DESCRIPTION} file are installed.

Documentation

Function documentation is accessible using the standard R \code{?} convention (e.g. documentation for \code{make_idw_map} can be accessed using \code{?make_idw_map}).

?make_idw_map

Make IDW maps

Start by loading and inspecting the 2017 EBS and NBS yellowfin sole CPUE data that is included with the package.

# Load 2017 EBS and NBS yellowfin sole data
yfs2017 <- akgfmaps::YFS2017
head(yfs2017)

The data includes the species common name, year, station id, latitude, longitude, and catch-per-unit effort in kilograms per hectare. At minimum, the data must include latitude, longitude, and catch-per-unit effort data to plot CPUE.

The function for performing the IDW and making a plot is \code{make_idw_map}. The function uses gstat to perform the IDW, then raster, stars, and sf for plotting. As arguments to the \code{make_idw_map} function, Users can specify:

• The domain for IDW and mapping, using the 'region' argument. Currently, maps can automatically be generated standard EBS shelf survey area (region = "bs.south") and full EBS shelf (region = "bs.all").
• Set break points for plotting, manually or algorithmically, using the 'set.breaks' argument. This can be a numeric vector of breaks (manual) or a character vector passed to \code{classInt::classIntervals) to algorithmically define breaks.
• Generate a custom extrpolation domain for the output. This is automatically generated for "bs.south" and "bs.all"
• The resolution of the extrapolation grid using the 'grid.cell' argument. Extrpolation takes longer for riner resolution grids.
• Add a key.title (e.g. species name).
• Define an input coordinate reference system. Retain the default setting if input is in degrees latitude and longitude.
• Define an output coordinate reference system that will be used for plotting (defaults to Albers Equal Area).
• Set the number of 'neighbor' points used for IDW. Defaults to 4, more neighbors takes longer to run.
• Data can be passed to the function using a data.frame or vectors. If using a data.frame, column names need to include CPUE_KGHA, LATITUDE, and LONGITUDE.

See function documentation for default arguments.

Below, we call the make_idw_map argument, specifying: * region = "bs.all" (northern + southern EBS), which has a predefined extrapolation area. * \code{grid.cell = c(0.02,0.02)

opt1 <- make_idw_map(x = yfs2017, # Pass data as a data frame
             region = "bs.all", # Predefined bs.all area
             set.breaks = "jenks", # Gets Jenks breaks from classint::classIntervals()
             in.crs = "+proj=longlat", # Set input coordinate reference system
             out.crs = "EPSG:3338", # Set output coordinate reference system
             grid.cell = c(20000, 20000), # 20x20km grid
             key.title = "yellowfin sole") # Include yellowfin sole in the legend title

The \code{make_idw_map} function returns a list containing object which contain information about the region, number of breaks, and key.title, which are used by other functions to manipulate plots and add labels.

opt1$region
opt1$n.breaks
opt1$key.title

The extrapolation.grid is a stars object that contains discrete CPUE values.

class(opt1$extrapolation.grid)

The continuous.grid is a stars object that contains continuous CPUE values.

class(opt1$continuous.grid)

The last object in the list is a ggplot object.

opt1$plot

Map layers are included in the package and do not need to be provided separately. Changes to text sizing, color palette, and addition of plot labels is done using change_text_size, change_fill_color, and add_idw

Changing colors

The \code{change_fill_color} function changes the fill color palette using predefined color palettes. Refer to the function documentation for a list of colors. Below, I show two transformations, one to grey scale and one to red, using pipe operators \code{|>} and passing the new fill color to the \code{new.scheme} argument. If no argument is provided to new.scheme, a color palette selector tool will open in a separate window \code{colorspace::choose_palette()}. The predefined color palettes are colorblind and grayscale friendly.

# Grey Note that the WHOLE list that was returned by make_idw_map should be passed to the function.
opt1.grey <- opt1 |> change_fill_color(new.scheme = "grey", show.plot = FALSE)

# Red
opt1.red <- opt1 |> change_fill_color(new.scheme = "red", show.plot = FALSE)

opt1.grey$plot
opt1.red$plot

The returned list objects retain the same structure as the objects from make_idw_map

Adding labels

Labels can be added by passing a data frame of labels, locations, and label types to the \code{add_idw_labels} function. Alternatively, pre-defined labels can be added to the plot based on the region. Below, pre-defined labels are added to the plot using \code{geom_text} and \code{geom_shadowtext}. Below, labels and a scale bar are added to the greyscale plot using pre-defined settings based on the plot region.

opt1.grey <- opt1.grey |> add_map_labels()
opt1.grey$plot

Custom labels can also be appended to the object, with sizing and orientation based on defined 'types.' In the example below, a data frame is passed with plotting locations. Internally, the \code{add_idw_plot_labels} function transforms long/lat coordinates in eastings to the Albers Equal Area projection of the map. See function documentation for options to turn on/off spatial transformations.

# ONLY custom label
opt1.red2 <- opt1.red |> add_map_labels(new.places = data.frame(type = "islands", region = "bs.all", x = -163, y = 57, lab = "Bristol Bay"), lab.replace = TRUE)
opt1.red2$plot

# Add custom label
opt1.red <- opt1.red |> add_map_labels(new.places = data.frame(type = "islands", region = "bs.all", x = -163, y = 57, lab = "Bristol Bay"), lab.replace = FALSE)
opt1.red$plot

Create map file

After labels and colors are added to the map, the \code{create_map_file()} function can be used to create publication-quality png files. It automatically rescales axis and text labels for some pre-defined plot sizes. Output file names can be specified using file.prefix, or they will be extracted automatically from the input list. The function passes plotting arguments to \code{png()}. The argument \code{try.change_text_size} indicates whether the function should try to resize text based on a pre-defined scaling factor.

opt1.grey |>
create_map_file(width = 8, # Passed to png()
                height = 6,  # Passed to png() 
                units = "in",  # Passed to png()
                res = 300,  # Passed to png()
                bg = "transparent", # Passed to png()
                file.path = "./", 
                try.change_text_size = TRUE)

Resizing text

The \code{change_text_size()} function can be used resize plot text prior to creating a map if the desired size for the map is not pre-defined.

opt.grey2 <- opt1.grey |> change_text_size(size.mult = 0.5)
opt.grey2$plot

Exploring break options

See the breaks vignette for details.

afsc-gap-products/akgfmaps documentation built on April 14, 2025, 7:13 p.m.