In mabecker89/wildRtrax: Environmental sensor data management and analytics to and from WildTrax

The recommended workflow to wrangle together data for analysis in wildRtrax is as follows. Once you have your data from wt_download_report().

• Remove unneeded species: wt_tidy_species()
• Convert TMTT counts to numeric: wt_replace_tmtt()
• Convert the species to a wide format: wt_make_wide()
• Calculate statistical offsets to account for imperfection detection following the QPAD method: wt_qpad_offsets()

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

load("package.RData")

# Attach package
library(wildRtrax)
library(tidyverse)

# Start by getting everything you need
Sys.setenv(WT_USERNAME = 'guest', WT_PASSWORD = 'Apple123')
wt_auth()
my_report <- wt_download_report(project_id = 605, sensor_id = 'ARU', reports = "main", weather_cols = F) %>%
  tibble::as_tibble()

Data wrangling

Now let's start cleaning things up a bit. We don't need certain abiotic codes so let's remove those with wt_tidy_species(),

my_tidy_data <- wt_tidy_species(my_report, remove = "mammal", zerofill=F)

# Difference in rows
round((nrow(my_tidy_data)/nrow(my_report)),2)

So about 5% of detections were mammals. Next, convert TMTT counts to numeric:

my_tmtt_data <- wt_replace_tmtt(data = my_tidy_data, calc = "round")

and finally widen the data into a species matrix.

my_wide_data <- wt_make_wide(data = my_tmtt_data, sound = "all")

head(my_wide_data)

Offsets

Now you can calculate statistical offsets to account for imperfection detection following the QPAD method.

my_offset_data <- wt_qpad_offsets(data = my_wide_data, species = "all", version = 3, together = TRUE)

head(my_offset_data)

Occupancy modelling

You can also perform a single-season, single-species occupancy work flow using wt_format_occupancy() once the data is downloaded.

dat.occu <- wt_format_occupancy(my_report, species="WTSP", siteCovs=NULL)
mod <- unmarked::occu(~ 1 ~ 1, dat.occu)
mod

mabecker89/wildRtrax documentation built on Feb. 5, 2024, 8:50 a.m.