In pratikunterwegs/watlastools: Tools for High Resolution Tracking Data from WATLAS

This vignette shows the basics of using watlastools. It is a work in progress. watlastools has two main purposes: accessing and cleaning data, and identifying residence patches from the data. Only the latter is covered here.

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(watlastools)

Making residence patches

Residence patches are the end goal of the watlastools workflow. They are a conveninent way of synthesising large amounts of tracking data into a few spatially explicit units that have a set of associated (mean) coordinates, as well as ecologically relevant coordinates such as duration.

Begin by reading in the data.

# read data and see leading lines
library(data.table)
data <- fread("vignette_data/413_004_revisit.csv")

head(data)

Now make residence patches after inferring residence from gaps in the data.

# run the three main functions on the data
# first infer residence
temp_data <- wat_infer_residence(data = data,
                                 inf_patch_time_diff = 30,
                                 inf_patch_spat_diff = 100)

# classify residence points
temp_data <- wat_classify_points(data = temp_data,
                                 lim_res_time = 2)

# make residence patches with sensible parameters
patches <- wat_make_res_patch(data = temp_data,
                              buffer_radius = 10,
                              lim_spat_indep = 100,
                              lim_time_indep = 30,
                              lim_rest_indep = 30,
                              min_fixes = 3)

Getting data from patches

Get and view the different kinds of patch data

Patch summary

The patch summary is a data.table and has no complex (list, sf) column types. The patches can be visualised using ggplot (or anything else).

# get patch summary
patch_summary <- wat_get_patch_summary(res_patch_data = patches,
                                       which_data = "summary")

# get the class of the full object and the column types
class(patch_summary)

Patch points

The patch points are the raw points that make up the residence patches.

# get patch points
patch_points <- wat_get_patch_summary(res_patch_data = patches,
                                      which_data = "points")

# get class and column type
class(patch_points)

Patch polygons

This should have the added class sf, and the polygons column should have the type sfc_MULTIPOLYGON.

# get patch points
patch_polygons <- wat_get_patch_summary(res_patch_data = patches,
                                      which_data = "spatial")

# get class and column type
class(patch_polygons)

Plot data types

Look at patch points and summary. Polygons are constructed as x metre buffers around patch points, and are not covered here. The code to make this plot is in the watlas_vignette.Rmd fine in the package's vignettes folder (the source for this document).

# plotting libraries
library(ggplot2)
library(patchwork)

# plot patch summary
fig_patch_summary <- ggplot(patch_summary, aes(x_mean, y_mean))+
  geom_path(col = "black", lty = 2, lwd = 0.2)+
  geom_point(aes(size = duration/60,
                 col = waterlevel_start),
             stroke = 2,
             shape = 21, fill = "black",
             alpha = 0.9)+
  geom_text(aes(label = patch), 
            col = "white", size = 2)+

  scale_size(range = c(3,10))+
  scale_colour_distiller(palette = "RdBu", direction = 1)+

  scale_x_continuous(expand = expansion(mult = c(0.1)))+
  scale_y_continuous(expand = expansion(mult = c(0.1)))+

  coord_sf(crs = 32631)+
  # annotation_scale(location = "br", width_hint = 0.1, text_cex = 1)+

  theme_bw()+
  theme(legend.position = "bottom",
        axis.text = element_blank(),
        axis.ticks = element_blank())+

  labs(x = NULL, y = NULL, 
       col = "waterlevel (cm NAP)", size = "residence (mins)",
       title = "(b) Patch summary data")

# plot patch points
fig_patch_points <- ggplot(patch_points, aes(x, y))+
  geom_path(col = "black", lty = 2, lwd = 0.1)+
  geom_point(aes(col = as.factor(patch)),
             shape = 21, alpha = 0.5,
                      show.legend = F)+

  scale_x_continuous(expand = expansion(mult = c(0.1)))+
  scale_y_continuous(expand = expansion(mult = c(0.1)))+
  scale_colour_manual(values = rep(c('red', 'blue'), 
                                   (length(unique(patch_points$patch))+1) /2))+
  coord_sf(crs = 32631)+
  # annotation_scale(location = "br", width_hint = 0.1, text_cex = 1)+

  theme_bw()+
  theme(legend.position = "bottom",
        axis.text = element_blank(),
        axis.ticks = element_blank())+

  labs(x = NULL, y = NULL,
       title = "(a) Patch point data: tag 413, tidal cycle 04")

library(patchwork)

fig_patch_points / fig_patch_summary

pratikunterwegs/watlastools documentation built on Nov. 5, 2021, 2:07 p.m.