window

This project introduces a sliding window analysis function for the ctmm R package.

Features

• Custom time-series object class for sliding window analysis of animal tracking data
• Extracts point estimates and confidence intervals for selected variable estimates
• Customizable window size and time step options
• Flexible functionality for individual and population level estimates
• Compatible with the ctmm package's existing framework
• Plot point estimats, confidence intervals and (optional) covariates to visualize correlation

Getting Started

Prerequisites

• R
• The ctmm package installed.

Installation

Clone the repository and set up the environment:

git clone https://github.com/ctmm-initiative/window.git
cd window

Arguments

• data: tracking data of ctmm class
• CTMM: Guess model used as a starting point for the model fitting process. Use ctmm.guess(data)
• variable: calculates parameter estimate : "area", "diffusion", "speed, "velocity", "position"
• dt.min: minimum time step between Time series windows as difftime object
• window: window size as difftime object
• max_windows: Option for setting a limit to window estimates calculated for the time series (TS)
• select: More rigorous method for selecting movement model by default. For faster modeling fitting make select = FALSE
• recycle: Option to use the previous model fit as a starting point for the next model fitting process in the timeseries
• Guassian: Option when estimating speed parameter (Default set to FALSE)
• covariate: Option for including covariate in the TS object class and plotting along parameter estimate (name of column)
• release: Option to examine trends in terms of days since release or days since first track (Default set to FALSE)

Usage

Create an object of class TS using the animal tracking data, and then use plot() to visualize the TS object

library(ctmm)

  # Download example data
data(buffalo)

  # Individual example dataset
individual_data <- buffalo$Cilla
individual_GUESS <- ctmm.guess(individual_data, interactive = FALSE)

  # Population example dataset
population_data <- buffalo[c(1, 3, 6)]
population_GUESS <- lapply(population_data, function(population_data) ctmm.guess(population_data, interactive = FALSE))

  # Arguments
min_time_step <- as.difftime(10, units = "days")
window <- as.difftime(30, units = "days")

  # Create TS of individual window estimates
individual_TS <- slide(data = individual_data, CTMM = individual_GUESS, window = window, dt.min = min_time_step, recycle = TRUE)

  # Create TS of population window estimates
population_TS <- slide(data = population_data, CTMM = population_GUESS, window = window, dt.min = min_time_step, recycle = TRUE)

  # plot results
plot(population_TS)
plot(individual_TS)

For a more detailed guide, view example_usage.R located in the source-code folder.

Acknowledgments

• This project was developed by Michael Garan under the guidance of Dr. Christen Fleming.
• Maintained and edited by Nozomu Hirama.

Contact

For questions or feedback, contact:

Michael Garan
Email: michael.d.garan@gmail.com

Nozomu Hirama
Email: nozomu.hirama@ucf.edu