Perceptual_tSNE

Code to make perceptual embedding plots

With Docker- implemented by Tsan-Kuang Lee

Installation

Install docker and docker-compose on the host machine.
git clone https://github.com/DrewWham/Perceptual_tSNE.git

Run

Enter the repo directory cd Perceptual_tSNE
Edit command.sh as needed (see Usage below)

./run.sh

This will take a while while it installs all R packages and Python packages. The subsequent runs will take shorter.

Without Docker

Dependencies

Python 3.5 or later
R 3.5

Quick Start

-open a terminal and git clone this repo

-then change your working directory to the repo with: cd Perceptual_tSNE

-install the required python 3 packages with: pip install -r ./required/requirements.txt

-install the required R packages with: Rscript ./required/requirements.R

-run python Perceptual_tSNE.py -h to view help and test that all dependencies are installed correctly.

Usage

run python Perceptual_tSNE.py from the directory it is located.

The following options are required:
-i specifies the name & location of the input image directory
-n specifies the base name to use for program outputs -p specifies the perplexity parameter to be used in t-SNE analysis -g allows the use of a GPU, pass -use_gpu TRUE if you have a CUDA enabled GPU available on the system with gpu relevent dependencies installed.

The following options are available:
-s specifies the random seed to use for the random number generator.

To produce the example output files run:

python Perceptual_tSNE.py -i "./input_images/females" -n females_1 -s 1 -p 30 -g FALSE

Please credit:
Our methods paper for the use of this repo and the application of this technique for quantifying perceptual distance in biological systems:Measuring Perceptual Distance of Organismal Color Pattern using the Features of Deep Neural Networks. (2019) Drew C Wham, Briana D Ezray, Heather M Hines. https://doi.org/10.1101/736306

Our application paper for the use of this technique in quantifying bumblebee mimicry: Müllerian mimicry in bumble bees is a transient continuum. (2019) Briana D. Ezray, Drew C. Wham, Carrie Hill, Heather M. Hines https://doi.org/10.1101/513275

Zhang et al. (2018)'s paper for the base machine learning technique for quantifying perceptual distance of images: https://github.com/richzhang/PerceptualSimilarity https://richzhang.github.io/PerceptualSimilarity/ Zhang et al. 2018 for perceptual similarity code.

DrewWham/Perceptual_tSNE