CartogramPy

CartogramPy is meant for simplifying making Gastner-Newman cartograms using python.

^^ sample output as density grows inside the green bordered regions

Getting Started

to generate a cartogram, you need to give the generate.cartogram function a 2d input variable matrix, and an image to transform. something like generate.cartogram(im, z) where im is your image, and z is your variable matrix that govern the cartogram.

here is a more in depth example:

import numpy as np                                                              
from PIL import Image, ImageDraw                                                
import generate                                                                 
                                                                                
# the image that you will distort into a cartogram                              
#im = Image.open('path/to/image.png')                                           
# for simple test, just use numpy array                                         
w,h = 500,500                                                                   
im = Image.fromarray(np.zeros((w,h)), mode='RGB')                               
d = ImageDraw.Draw(im)                                                          
                                                                                
# the data that will determine the distortion- if you were doing population, say, the points of this 2d a
# the population of whichever region that point falls in.                       
z = np.zeros((w, h)) # or whatever                                              
z += 1 # add a baseline... this seems to help cartograms from distorting        
# as an example, add some square regions with higher "density"                  
squares = [                                                                     
    [100,100,200,200],                                                          
    [250,250,400,400],                                                          
    [100,400,200,450],                                                          
]                                                                               
for square in squares:                                                          
    x1, y1, x2, y2 = square                                                     
    # draw boxes in green. these will get distorted by the transformation                                
    d.line([(x1,y1), (x1,y2), (x2,y2), (x2,y1), (x1,y1)], fill=(0,255,0)) # outline them
    z[y1:y2,x1:x2] += 3 # density is 4 times the baseline of 1                  
im = generate.cartogram(im, z)                                                  
d = ImageDraw.Draw(im)                                                          
# show initial boxes in red for referrence                                      
for square in squares:                                                          
    x1, y1, x2, y2 = square                                                     
    d.line([(x1,y1), (x1,y2), (x2,y2), (x2,y1), (x1,y1)], fill=(255,0,0))       
im.show()

Prerequisites

This little library requires the command line tool cart. It expects that cart command in bash will work. Follow the instructions on cart's page here for how to install it, as well as interp, which comes with it when you download it.

you will also need some good old python library standbys: numpy and PIL

Contributing

ill look at any pull request.

Author

Patrick Brooks

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments

Hat tip to cart author
accidentally developed as a filed attempt to help map carbon emissions

kittles / cartogrampy