Read Me Text File to accompany supplementary material for: 
Variation in carotenoid-containing retinal oil droplets correlates with variation in perception of carotenoid coloration
Eleanor M. Caves, Lorian E. Schweikert, Patrick A. Green, Carlos Taboada, Matthew N. Zipple, Susan Peters, Stephen Nowicki, and Sönke Johnsen

ManipAnalysis.R
This file contains the R code necessary to generate all of the linear mixed models described in the text. It requires the user to upload "Upload_File.csv" and "lambda2.csv"

"Upload_file"
This file contains the raw data necessary to generate all of the linear mixed models described in the text. 
Columns are as follows:
bird.ID=an individual identifier for each bird used in the study
col 1= for a given color discrimination task, the first color in the color pair; values range from 1 (most red) to 8 (most orange)
col 2= for the same color discrimination task, the second color in the color pair; values range from 1 (most red) to 8 (most orange)
pass= out of 7 trials for a given color combination, the proportion passed by a given bird
n = number of trials in which the bird participated
across 1 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 1; 1 = yes, 0 = no
across 2 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 2; 1 = yes, 0 = no
across 3 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 3; 1 = yes, 0 = no
across 4 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 4; 1 = yes, 0 = no
across 5 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 5; 1 = yes, 0 = no
across 6 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 6; 1 = yes, 0 = no
across 7 = a binomial indicator of whether the color combination of col1 and col2 crossed color step 7; 1 = yes, 0 = no
treatment= the treatment group for the bird: control, carotenoid minus or carotenoid replaced
lambda = the mean lambda mid value for the 15 R-type oil droplets measured from a given bird
lambcent = the mean-centered lambda mid value for the 15 R-type oil droplets measured from a given bird; 0 = population average
col 1x1 = From the RNL model; The X1 coordinate representing color 1
col 1x2 = From the RNL model; The X2 coordinate representing color 1
col 2x1 = From the RNL model; The X1 coordinate representing color 2
col 2x2 = From the RNL model; The X2 coordinate representing color 2
col1brightness = the double cone quantum catch from color 1
col2brightness = the double cone quantum catch fro color 2
edist = the euclidean distance in chromaticity space between color 1 and color 2
contrast = michelson contrast, i.e. the ratio of the sum to the difference of brightness between colors 1 and 2
age = age of the bird (in days) at the start of trials

"lambda2.csv"
This file contains the average lambda-mid values for each bird, as well as the random effect coefficient estimates of crossing the 5-6 color step on color discrimination for each bird.
Columns are as follows:
Treatment = the treatment group for the bird: control, carotenoid minus or carotenoid replaced
Bird.ID=an individual identifier for each bird used in the study
lambda = the mean lambda mid value for the 15 R-type oil droplets measured from a given bird
across5 = the random effect coefficient estimates of crossing the 5-6 color step on color discrimination for each bird. Data is missing for birds that did not successfully complete color discrimination trials.

MSP_Analysis_Code_Supplement.Rmd
An R-Markdown file containing the code necessary to fit logistic models to oil droplet transmittance spectra and extract lambda mid values.
This code can be run on a sample file "Sample_oil droplet transmission file_supplement.txt"
Raw oil droplet transmittance files for all measured droplets available upon request. 

lambda_mid_repeatability.csv and lambda_mid_repeatability_R.R
These files provide the data and R code used to calculate repeatability of lambda mid. 

*Note that all analyses were originally conducted in R version 3.5.0