synthetic spectral reflectance functions and basic colour naming data

Each of pcaPlusReflectances.csv, fourierPlusReflectances.csv and sigmoidSumReflectances.csv contain 100,000 randomly generated spectral reflectance models according to the model specified in the filename. See paper ('Reconciling the statistics of spectral reflectance and colour' expected to be published in PLOS One) for a description of models. First line of each file are the wavelengths (nm) at which the functions are sampled, each subsequent line is one function.

The file fuzzyBCC.csv contains membership values for the basic colour categories sampled over the RGB cube. See paper above for some details, and https://doi.org/10.1371/journal.pone.0216296 for more. Sampling of the RGB cube is at 32*32*32 resolution. Index 1 maps to a minimum (0) RGB value, and index 32 to a maximum (255) RGB value; other indices map to equally spaced RGB values within these limits. Each row of the file gives R,G,B values for the sample, followed by the membership probabilities for the BCCs as indicated in the 1st row of the file. The membership values in each row have unit sum.