Evidence of self-care tooling and phylogenetic modelling reveal parrot tool use is not rare

Using YouTube and phylogenetic modelling to identify tool use in parrots.

Getting started

Software requirements

This code has been tested on the following operating systems:

• Windows 10 Education 22H2 64-bit
• Linux 3.10.0-693.2.2.el7.x86_64 x86_64
• macOS Ventura 13.4

This code requires R (v4.2.1) and the following R package versions:

• ape (v5.6-2)
• brms (v2.16.0)
• cowplot (v1.1.1)
• dagitty (v0.3-1)
• future (v1.27.0)
• future.callr (v0.8.0)
• geiger (v2.0.10)
• ggdag (v0.2.6)
• ggridges (v0.5.3)
• ggstance (v0.3.6)
• magick (v2.7.3)
• papaja (v0.1.1.9001)
• phangorn (v2.9.0)
• phytools (v1.0-3)
• readxl (v1.4.1)
• reshape2 (v1.4.4)
• rethinking (v2.21)
• rstan (v2.26.13)
• scales (v1.2.1)
• sdamr (v0.2.0)
• tarchetypes (v0.7.1)
• targets (v0.13.5)
• tidybayes (v3.0.2)
• tidyverse (v1.3.2)

Installation guide

To run this code, you will need to install R and the following R packages:

install.packages(
    c("ape", "brms", "cowplot", "dagitty", "future",
      "future.callr", "geiger", "ggdag", "ggridges",
      "ggstance", "magick", "papaja", "phangorn", 
      "phytools", "readxl", "reshape2", "rstan", 
      "scales", "sdamr", "tarchetypes", "targets",
      "tidybayes", "tidyverse")
)

You will also require the ggtree package, which can be installed using the following code:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("ggtree")

Finally, you will require the rethinking package, which can be installed by following the instructions here or using the following code:

install.packages("cmdstanr")
cmdstanr::install_cmdstan()
install.packages(c("coda","mvtnorm","devtools","loo","dagitty","shape"))
devtools::install_github("rmcelreath/rethinking")

These installation steps should typically take no longer than half an hour on a normal desktop computer.

Run demo

A short demo executes the following steps from the full analysis pipeline: (1) loads the data, (2) estimates phylogenetic signal for tool use across 100 posterior samples from the parrot phylogeny, and (3) combines these estimates into a single posterior distribution.

To run this short demo:

1. Download this code repository to your local machine using git clone https://github.com/ScottClaessens/phyloParrot or by downloading the .zip file from GitHub
2. Set the working directory to this code repository setwd("myPath/phyloParrot") on your local machine
3. Load the targets package with library(targets)
4. Run tar_make(phySignal1) in the command line
5. To load the model object, run tar_load(phySignal1) in the command line

This demo should typically take 5-10 minutes on a normal desktop computer.

Run full analysis pipeline

The full analysis pipeline can be viewed using targets::tar_visnetwork(). Running this pipeline will reproduce all figures, tables, and quantitative results from the manuscript, as well as a PDF of the manuscript itself.

To run the full analysis pipeline:

1. Download this code repository to your local machine using git clone https://github.com/ScottClaessens/phyloParrot or by downloading the .zip file from GitHub
2. Set the working directory to this code repository setwd("myPath/phyloParrot") on your local machine
3. Load the targets package with library(targets)
4. To run all analyses, run tar_make() in the command line
5. To load individual targets into your environment, run tar_load(targetName) in the command line

Note that some of the models, especially for the cross-validation, will take a long time to run. The full analysis pipeline will take several days to run in series. You can run the pipeline in parallel using tar_make_clustermq() or tar_make_future() (see here).

Help

Any issues, please email scott.claessens@gmail.com.

Authors

Scott Claessens, scott.claessens@gmail.com