🎨 Theory Of Colors

Functionalities to work with colors

Theory of Colours (German: Zur Farbenlehre) is a book by Johann Wolfgang von Goethe about the poet's views on the nature of colours and how these are perceived by humans – Wikipedia

➡️ Functionalities

• ColorCast: Cast colors
• Nearest color:
  • NearestColorViaEmbedding: Find nearest color name (from a reference list) to any possible color string (using text embedding similarity)
  • NearestColorViaColorSpace: Find nearest color name to any possible (RGB, HSV, HEX) color value (using euclidian distance on RGB or HSV color space)
• ColorComplement: Return opposite color

1️⃣ Convert colors from/to RGB (0-1 floats), RGB (0-255 int) HEX, HSV and strings

from theoryofcolors import ColorCast

cc = ColorCast()

print(cc("#FF0000", from_="hex", to_="hsv"))
print(cc((1.0,0.,0.), "rgb", "rgbint"))
print(cc("red", "string", "rgb"))

2️⃣ 🅰️ Find nearest color to given string – using Text Embedding on color string name

• Similarity is calculated based on text embedding similarity (cosine similarity) between given word a prefined reference list of color names.
• model_name: You can use any model from sentence-transformers. Default is all-distilroberta-v1.
• reference_colors: You can choose between four diffrent reference color lists:
  • 8 base colors
  • 949 xkcd colors
  • 148 css4 colors
  • all colors (1048 colors in total)

from theoryofcolors import NearestColorViaEmbedding

nc = NearestColorViaEmbedding()

# Nearest output as string
print(nc(color_string="dark yellow")) # dark yellow
print(nc("fire red"))                 # fire engine red
print(nc("tocxic greeeen"))           # toxic green

# Nearest output as colorized string (works in terminal and jupyter notebook)
nc.print("fire red")
nc.print("brightblue") # lightblue

# Use an other embedding model
nc = NearestColorViaEmbedding(model_name='all-mpnet-base-v2')
nc.print("pink", arrow=True)
nc.print("orange", arrow=True)
nc.print("darkblue", arrow=True)

# Use other reference colors (match to 8 base colors only)
nc = NearestColorViaEmbedding(reference_colors="base")
nc.print("pink", arrow=True) # pink => magenta
nc.print("orange", arrow=True) # orange => green
nc.print("darkblue", arrow=True) # darkblue => blue

2️⃣ 🅱️ Find nearest color string to given RGB (or HSV, HEX) value – using euclidian distance on RGB or HSV similarity color space

from theoryofcolors import NearestColorViaColorSpace

color = (0.0,.9,.1)
from_ = "rgb" # you can also use "hsv", "hex"

nc = NearestColorViaColorSpace(reference_colors="all", similarity_color_space="rgb")
res = nc.nearest(color=color, from_=from_, to_="string", top_k=3)
res
#> ['vibrant green', 'neon green', 'bright green']
nc.print(color, from_=from_, arrow=True)
#> source color: (0.0, 0.9, 0.1) ====> target color: vibrant green, rgb: (0.04, 0.87, 0.03)


nc = NearestColorViaColorSpace(reference_colors="base", similarity_color_space="rgb")
res = nc.nearest(color=color, from_=from_,  to_="string",  top_k=3)
res
#> ['green', 'cyan', 'black']

3️⃣ Find complement color

from theoryofcolors import ColorComplement

comp = ColorComplement()

comp((1,0,0), from_="rgb", to_="hex")
comp((1,1,1), "rgb")
comp((.5,0,0), "rgb")

4️⃣ Misc

from theoryofcolors.misc import load_color_dict

# Map string to hex
color_mapper = load_color_dict("base") # 8 colors
color_mapper = load_color_dict("xkcd") # 949 colors
color_mapper = load_color_dict("css4") # 148 colors
color_mapper = load_color_dict("all")  # 1048 colors

ℹ️ Install

• Install from GitHub

pip install git+https://github.com/phiyodr/theoryofcolors

• Install from PyPI

pip install theoryofcolors

🌻