| Title: | A package to reconstruct sporadic passive tracking data and measure the bias and uncertainty incurred through the process |
|---|---|
| Description: | This package is the corresponding code for the Methods in Ecology and Evolution article, "A novel process to infer the reliability of ecological information derived from passive acoustic telemetry track reconstruction". The intended workflow is to simulate tracks inside a specified polygon, calculate quadrat counts inside grid cells, derive passive detection data from the simulated tracks using geolocations of listening stations, reconstruct the tracks using continuous-time correlated random walk models, re-route any spurious track intersections with a specified boundary, calculate quadrat counts of reconstructed tracks within grid cells, and compare quadrat counts between simulated and reconstructed tracks. This process is designed to be iterated to achieve statistical power. The user can then proceed with modeling the relationship between environmental variables and a desirable result (i.e., a good fit) to determine what environmental variables affect the odds of a good fit in grid cells. |
| Authors: | Mary E. Bowers [aut, cre] |
| Maintainer: | Mary E. Bowers <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.1.0 |
| Built: | 2024-10-30 06:48:38 UTC |
| Source: | https://github.com/mebowers5/sporeg |
Compare grid cell counts
comp_trks( sim_trks, stations, land_barrier, vis_graph, multi.grid, HSgrid, snap_tolerance )comp_trks( sim_trks, stations, land_barrier, vis_graph, multi.grid, HSgrid, snap_tolerance )
sim_trks |
Tracks to which modeled tracks should be compared |
stations |
A sf object comprised of receiver station locations with a buffer around them that represents the range of the receiver (polygons) |
land_barrier |
A sf polygon of a barrier object around which tracks should be re-routed |
vis_graph |
A visibility graph created from the barrier object @seealso pathroutr::prt_visgraph |
multi.grid |
boolean |
HSgrid |
When multi.grid = FALSE, a sf polygon grid; When multi.grid = TRUE, a list of sf polygon grids |
snap_tolerance |
The tolerance (in meters) at which an intersection between a station and a track should snap to the station centroid. It is recommended that the snap_tolerance be equal to the station buffer size. |
A data frame with counts and differences by grid cell ID ("gid")
The function allows you to calculate counts per grid cell in demonstrative modeled movement data.
cts(sg, df)cts(sg, df)
sg |
spatial grid created from |
df |
sf object of reconstructed, re-routed, optionally buffered tracks created from |
A simple feature object with counts associated with grid cell IDs "gid"
This function determines the distance to shore and the density of receivers for each grid cell.
d_shore_rcvs(km, study_site, land_barrier, epsg, sts_pts)d_shore_rcvs(km, study_site, land_barrier, epsg, sts_pts)
km |
grid cell resolution in km. one-sided length of grid cell, assumes desired grid cell is to be squared |
study_site |
simple feature polygon object that encompasses the entire study site |
land_barrier |
simple feature (multi)polygon object to route tracks around |
epsg |
epsg code for desired coordinated system transformation |
sts_pts |
a simple feature (multi)point object representing receiver locations |
A simple feature multipolygon object with information on distance to shore from grid cell center, receiver presence/absence, counts, and densities
This function provides summary statistics on the densities and counts of receivers in a grid.
den_rcvs(df)den_rcvs(df)
df |
A simple feature polygon object (a grid) |
A data frame object with minimum, mean, and maximum of receiver densities and counts in km^-2
# Apply den_rcvs to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) rcv_dens <- lapply(res, den_rcvs) rcv_dens <- data.table::rbindlist(rcv_dens, idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))# Apply den_rcvs to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) rcv_dens <- lapply(res, den_rcvs) rcv_dens <- data.table::rbindlist(rcv_dens, idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))
This function allows you to calculate mean depth per grid cell.
depth_data(HSgrid, depth)depth_data(HSgrid, depth)
HSgrid |
list object of multiple grids |
depth |
simple feature point object with |
A simple feature multipolygon object with mean depth per grid cell
Different cut off
dif_co(df, depth_limit)dif_co(df, depth_limit)
df |
data frame object consisting of results from iterative reconstruction process |
depth_limit |
a depth value in meters that represents the new depth cut off of interest |
a data frame object containing the percentage of grid cells that contained a good fit out of all grid cells that had an average depth (mean_depth) less than or equal to the depth_limit
# Apply dif_co to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) depth_limit <- 300 #[m] dif_depth <- lapply(res, dif_co) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))# Apply dif_co to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) depth_limit <- 300 #[m] dif_depth <- lapply(res, dif_co) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))
This function summarizes minimum, maximum, and mean statistics for depth (mean_depth), distance from shore (d_shore), receiver count (count), receiver presence/absence (p_a), and the percentage of grid cells missing receivers for all of the good fits (g_fit).
gd_fts(df)gd_fts(df)
df |
data frame object consisting of results from iterative reconstruction process |
a data frame object with summary statistics from grid cells that contained good fits
# Apply gd_fts to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) gd_fit_char <- lapply(res, gd_fts) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))# Apply gd_fts to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) gd_fit_char <- lapply(res, gd_fts) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))
This function calculates the odds ratios for each variable in your model.
get_or(model)get_or(model)
model |
A model object created from the |
A data frame object with a point estimate and lower and upper 95% confidence interval values associated with each covariate
# Get the odds ratio for each variable in the final model load(system.file("extdata", "fit2.100km.Rda", package = "sporeg")) odds_100km <- get_or(fit2.100km) %>% dplyr::mutate(res_name = "100km")# Get the odds ratio for each variable in the final model load(system.file("extdata", "fit2.100km.Rda", package = "sporeg")) odds_100km <- get_or(fit2.100km) %>% dplyr::mutate(res_name = "100km")
Gaps characteristics
gps_chr(df)gps_chr(df)
df |
data frame object consisting of results from iterative reconstruction process |
a data frame object with summary statistics that provide insight into the locations of the gaps in the network receiver array that were closed by the reconstruction process
# Apply gps_chr to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) clsd_gp_chars <- lapply(res, gps_chr) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))# Apply gps_chr to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) clsd_gp_chars <- lapply(res, gps_chr) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))
This function calculates how well the reconstructions closed gaps in the network receiver array
gps_fld(df)gps_fld(df)
df |
data frame object consisting of results from iterative reconstruction process |
a data frame object with a percentage of grid cells that contained a good fit out of those that lacked receivers
# Apply gps_fld to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) clsd_gps <- lapply(res, gps_fld) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))# Apply gps_fld to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) clsd_gps <- lapply(res, gps_fld) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))
This function creates a grid inside a specified polygon.
grid_res(km, study_site, epsg, what)grid_res(km, study_site, epsg, what)
km |
grid cell resolution in km. one-sided length of grid cell, assumes desired grid cell is to be squared |
study_site |
simple feature polygon object in which grid cells should be created |
epsg |
epsg numeric code for desired coordinate system transformation |
what |
|
grid comprised of polygons or points depending on what parameter
# NOTE: The study site must be in a projected coordinate system (e.g., WGS 84; EPSG: 3857) # when it is initially fed into the function grid_res library(sporeg) library(dplyr) library(sf) load(system.file("extdata", "site_depth.Rda", package = "sporeg")) site_depth <- site_depth %>% sf::st_transform(., 3857) HS_100km_grid <- grid_res(100, site_depth, 4269, "polygons")# NOTE: The study site must be in a projected coordinate system (e.g., WGS 84; EPSG: 3857) # when it is initially fed into the function grid_res library(sporeg) library(dplyr) library(sf) load(system.file("extdata", "site_depth.Rda", package = "sporeg")) site_depth <- site_depth %>% sf::st_transform(., 3857) HS_100km_grid <- grid_res(100, site_depth, 4269, "polygons")
This functions creates a line by connecting start and end points.
make_line(start_x, start_y, end_x, end_y)make_line(start_x, start_y, end_x, end_y)
start_x |
longitudinal sf coordinate object from the desired start point |
start_y |
latitudinal sf coordinate object from the desired start point |
end_x |
longitudinal sf coordinate object from the desired end point |
end_y |
latitudinal sf coordinate object from the desired end point |
A simple feature geometry object or simple feature linestring object
library(sporeg) library(dplyr) library(sf) library(purrr) library(tidyr) load(system.file("extdata", "at_dly_locs.Rda", package = "sporeg")) at_lines <- at_dly_locs %>% dplyr::group_by(ID, time) %>% sf::st_transform(3857) %>% dplyr::summarise(pt = sf::st_combine(geometry)) %>% sf::st_centroid() %>% dplyr::mutate(lat = sf::st_coordinates(pt)[,2], lon = sf::st_coordinates(pt)[,1]) %>% dplyr::arrange(ID, time) %>% #Order data for making lines dplyr::mutate(start_x = lon, start_y = lat, end_x = dplyr::lead(lon), end_y = dplyr::lead(lat)) %>% sf::st_as_sf(coords = c("lon", "lat"), crs = 3857) %>% dplyr::filter(!is.na(end_y)) %>% tidyr::nest() %>% dplyr::mutate( data = purrr::map(data, ~ dplyr::mutate(.x, x = purrr::pmap(.l = list(start_x, start_y, end_x, end_y), .f = make_line))))library(sporeg) library(dplyr) library(sf) library(purrr) library(tidyr) load(system.file("extdata", "at_dly_locs.Rda", package = "sporeg")) at_lines <- at_dly_locs %>% dplyr::group_by(ID, time) %>% sf::st_transform(3857) %>% dplyr::summarise(pt = sf::st_combine(geometry)) %>% sf::st_centroid() %>% dplyr::mutate(lat = sf::st_coordinates(pt)[,2], lon = sf::st_coordinates(pt)[,1]) %>% dplyr::arrange(ID, time) %>% #Order data for making lines dplyr::mutate(start_x = lon, start_y = lat, end_x = dplyr::lead(lon), end_y = dplyr::lead(lat)) %>% sf::st_as_sf(coords = c("lon", "lat"), crs = 3857) %>% dplyr::filter(!is.na(end_y)) %>% tidyr::nest() %>% dplyr::mutate( data = purrr::map(data, ~ dplyr::mutate(.x, x = purrr::pmap(.l = list(start_x, start_y, end_x, end_y), .f = make_line))))
This is a wrapper function for running a power analysis on iterative simulation and reconstruction methods process.
powr(output, sig.level, power, delta, n)powr(output, sig.level, power, delta, n)
output |
the resulting data frame from the iterative methods process |
sig.level |
numeric. the desired level of significance to achieve |
power |
numeric. the desired level of power to achieve |
delta |
numeric. the desired effect size to achieve |
n |
integer. the number of replicates/sample size. should be assigned |
A data frame object
# Use powr wrapper function on example results library(sporeg) library(dplyr) library(data.table) load(system.file("extdata", "results.Rda", package = "sporeg")) results <- lapply(results, data.table::rbindlist, idcol = 'resolution') results <- data.table::rbindlist(results, idcol = 'iteration') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") df_var <- zero_var(results) pow_stat <- df_var %>% dplyr::group_by(res_name, gid) %>% dplyr::summarise(sd = sd(dif)) %>% dplyr::ungroup() %>% dplyr::group_by(res_name) res <- pow_stat %>% dplyr::filter(res_name == "100km") anims <- 30 power <- 0.8 delta <- anims*0.01 # a delta within 1% of the total number of animals sig.level <- 0.95 n <- NULL pwr_100km <- powr(res, sig.level, power, delta) print("Grid cell resolution: 100 km x 100 km") pwr_100km# Use powr wrapper function on example results library(sporeg) library(dplyr) library(data.table) load(system.file("extdata", "results.Rda", package = "sporeg")) results <- lapply(results, data.table::rbindlist, idcol = 'resolution') results <- data.table::rbindlist(results, idcol = 'iteration') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") df_var <- zero_var(results) pow_stat <- df_var %>% dplyr::group_by(res_name, gid) %>% dplyr::summarise(sd = sd(dif)) %>% dplyr::ungroup() %>% dplyr::group_by(res_name) res <- pow_stat %>% dplyr::filter(res_name == "100km") anims <- 30 power <- 0.8 delta <- anims*0.01 # a delta within 1% of the total number of animals sig.level <- 0.95 n <- NULL pwr_100km <- powr(res, sig.level, power, delta) print("Grid cell resolution: 100 km x 100 km") pwr_100km
Receiver characteristics
rcv_chr(df)rcv_chr(df)
df |
data frame object consisting of results from iterative reconstruction process |
a data frame object with summary statistics that provide insight into the locations of the receivers in the network receiver array
# Apply rcv_chr to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) rcv_chars <- lapply(res, rcv_chr) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))# Apply rcv_chr to list of grid resolutions library(sporeg) library(dplyr) library(sf) library(data.table) load(system.file("extdata", "res.Rda", package = "sporeg")) rcv_chars <- lapply(res, rcv_chr) %>% data.table::rbindlist(., idcol = 'resolution') %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") %>% dplyr::mutate(res_name = ordered(res_name, levels = c("100km", "50km", "25km", "10km")))
This function simulates tracks inside a specified polygon.
simul_trks( anims, study_site, theta, vmin, vmax, rel_site, crs, n_days, initHeading )simul_trks( anims, study_site, theta, vmin, vmax, rel_site, crs, n_days, initHeading )
anims |
integer. quantity of desired animals to simulate |
study_site |
simple feature polygon object that encompasses area where tracks are allowed to be simulated. |
theta |
argument from |
vmin |
numeric. minimum velocity from which to sample step length |
vmax |
numeric. maximum velocity from which to sample step length |
rel_site |
simple feature polygon object in which simulated animals are "released" or where simulated tracks begin |
crs |
EPSG code for study_site polygon. Must be projected coordinate system |
n_days |
integer. number of days that tracks should be simulated |
initHeading |
argument from |
simple feature (multi)linestring object that represents individual simulated tracks
library(sporeg) library(sf) library(glatos) load(system.file("extdata", "rel_site.Rda", package = "sporeg")) load(system.file("extdata", "study_site.Rda", package = "sporeg")) anims <- 30 yr <- 1 theta <- c(0, 1.74) vmin <- 0.98 vmax <- 1.58 crs <- 3857 n_days <- 365*yr initHeading <- 0 tracks <- simul_trks(anims, study_site, theta, vmin, vmax, rel_site, crs, n_days, initHeading)library(sporeg) library(sf) library(glatos) load(system.file("extdata", "rel_site.Rda", package = "sporeg")) load(system.file("extdata", "study_site.Rda", package = "sporeg")) anims <- 30 yr <- 1 theta <- c(0, 1.74) vmin <- 0.98 vmax <- 1.58 crs <- 3857 n_days <- 365*yr initHeading <- 0 tracks <- simul_trks(anims, study_site, theta, vmin, vmax, rel_site, crs, n_days, initHeading)
This function allows you to re-route demonstrative movement data that has been reconstructed using a movement model around a polygon barrier.
sub_rrt(track_data, CRS, barrier, vis_graph, buffer)sub_rrt(track_data, CRS, barrier, vis_graph, buffer)
track_data |
point location data with latitude and longitude information |
CRS |
epsg code for desired coordinate system transformation |
barrier |
polygon or multipolygon sf object |
vis_graph |
vis_graph object created from |
buffer |
desired buffer size for re-routed tracks; generally should relate to range of receivers; distance should use the same units as the final coordinate system |
A simple feature (multi)polygon object
library(sporeg) library(dplyr) library(pathroutr) library(sf) load(system.file("extdata", "subset.Rda", package = "sporeg")) load(system.file("extdata", "atlcoast.Rda", package = "sporeg")) CRS <- 3857 barrier <- atlcoast vis_graph <- pathroutr::prt_visgraph(barrier) buffer <- 650 tbuff650 <- sub_rrt(subset, CRS, barrier, vis_graph, buffer)library(sporeg) library(dplyr) library(pathroutr) library(sf) load(system.file("extdata", "subset.Rda", package = "sporeg")) load(system.file("extdata", "atlcoast.Rda", package = "sporeg")) CRS <- 3857 barrier <- atlcoast vis_graph <- pathroutr::prt_visgraph(barrier) buffer <- 650 tbuff650 <- sub_rrt(subset, CRS, barrier, vis_graph, buffer)
This function removes grid cells that lacked any variance.
zero_var(df)zero_var(df)
df |
data frame object of iterative methods results |
a data frame object with grid IDs that did not exhibit zero variance
# Remove grid cells with zero variance library(sporeg) library(dplyr) load(system.file("extdata", "results.Rda", package = "sporeg")) results <- lapply(results, data.table::rbindlist, idcol = 'resolution') results <- data.table::rbindlist(results, idcol = 'iteration') results <- results %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") df_var <- zero_var(results)# Remove grid cells with zero variance library(sporeg) library(dplyr) load(system.file("extdata", "results.Rda", package = "sporeg")) results <- lapply(results, data.table::rbindlist, idcol = 'resolution') results <- data.table::rbindlist(results, idcol = 'iteration') results <- results %>% left_join(tibble(resolution = 1:4, res_name = c("100km", "50km", "25km", "10km")), by = "resolution") df_var <- zero_var(results)