A small Python script to generate random color sequences, e.g. for use in plotting. Just call the "generate_new_color(existing_colors,pastel_factor)" function to generate a random color that is (statistically) maximally different from all colors in "existing_colors". The "pastel_factor" parameter can be used to specify the "pasteliness"(?) of th…

generate_random_color.py

import random

def get_random_color(pastel_factor = 0.5):
    return [(x+pastel_factor)/(1.0+pastel_factor) for x in [random.uniform(0,1.0) for i in [1,2,3]]]

def color_distance(c1,c2):
    return sum([abs(x[0]-x[1]) for x in zip(c1,c2)])

def generate_new_color(existing_colors,pastel_factor = 0.5):
    max_distance = None
    best_color = None
    for i in range(0,100):
        color = get_random_color(pastel_factor = pastel_factor)
        if not existing_colors:
            return color
        best_distance = min([color_distance(color,c) for c in existing_colors])
        if not max_distance or best_distance > max_distance:
            max_distance = best_distance
            best_color = color
    return best_color

#Example:

if __name__ == '__main__':

  #To make your color choice reproducible, uncomment the following line:
  #random.seed(10)

  colors = []

  for i in range(0,10):
      colors.append(generate_new_color(colors),pastel_factor = 0.9)
    
  print "Your colors:",colors

@lifenautjoe

What kind of colors does this generate and how could we convert them to hex? 🤔

I believe you get RGB colors. I recommend python-colormath for conversion.

One thing to consider is that there is always the background colour of the plot which is very important a line should not be equal to!

Hi,
This solution works really well, much appreciated. I've taken the liberty to include it in my project concerned with analyzing cardiac action potentials, credited in the readme and in comments as The color generator script has been originally written by Andreas Dewes: https://gist.github.com/adewes/5884820

With a stable release on the horizon, I would like to ask if this is okay with you.

Hi @thomashastings, this code is public domain so you can freely reuse it, no attribution necessary!

Thank you for the clarification!