4
\$\begingroup\$

Given:

  • G is an arbitrary graph
  • partitions is a list of its edges divided into (in this case) 3 sets.

Color generation:

  • Its purpose is to generate one distinct color for each partition.
  • The function turns a number between 0 and 1 into an RGB color.
  • For the given example, with 3 partitions, it would be called with 1/3, 2/3, and 1, and return a list of codes for green, blue, and red.
  • For other numbers it would be used to spread the different colors as widely as possible.
  • Visually, think of a circle with R, G, and B equally spaced around it. The number is the fraction of the way around the circle from R to pick the color, so if the number is .4, the color would be between G and B, closer to the G.

Color assignment:

  • For each set in the partition, a different color is generated and assigned to a list of colors.
  • In this case edge_color_list would be 9 greens, 4 blues, and 2 reds.
  • The actual list is: [(0, 1.0, 0.0), (0.0, 0, 1.0), (0.0, 0, 1.0), (0, 1.0, 0.0), (0, 1.0, 0.0), (0, 1.0, 0.0), (0, 1.0, 0.0), (0, 1.0, 0.0), (0, 1.0, 0.0), (0, 1.0, 0.0), (0.0, 0, 1.0), (0, 1.0, 0.0), (1.0, 0, 0.0), (1.0, 0, 0.0), (0.0, 0, 1.0)]

Problem: The top and bottom sections can't really be changed, but the def_color() function and edge_color_list sections each look like they were written in C, and I think they could be done more elegantly.

It's obvious how I was thinking while writing it (i.e. how I would write it in C), but I'd like to know how python coders think.

I'm having trouble writing python without a heavy C accent (I wrote C for decades).

I understand the python language fairly well (or know how to recognize what I don't know and how to look it up).

And I've reached the stage where what I've written feels wrong.

But I still don't seem to think in native python:

#!/usr/bin/env python3
import matplotlib.pyplot as plot
import networkx as nx

G = nx.petersen_graph()
partitions = [{0, 3, 4, 5, 6, 7, 8, 9, 11}, {1, 2, 10, 14}, {12, 13}]

def get_color(fraction):
    if fraction < 1/3:
        color = (1-3*fraction, 3*fraction, 0)
    elif fraction < 2/3:
        fraction -= 1/3
        color = (0, 1-3*fraction, 3*fraction)
    else:
        fraction -= 2/3
        color = (3*fraction, 0, 1-3*fraction)
    return color

edge_color_list = [None] * len(G.edges())
for i in range(len(partitions)):
    items = list(partitions[i])   # convert from set
    for j in range(len(items)):
        edge_color_list[items[j]] = get_color((i+1)/len(partitions))

nx.draw(G, with_labels=True, pos=nx.circular_layout(G), width=2,
        node_color='pink', edge_color=edge_color_list)
plot.show()

Any suggestions about how I could have thought differently while writing the above would be appreciated.

\$\endgroup\$
2
  • 5
    \$\begingroup\$ Please tell us more about the purpose of the code. What does it do and what prompted you to write it? \$\endgroup\$
    – Mast
    Aug 7 '20 at 14:56
  • 1
    \$\begingroup\$ @Mast, I've added more details about the code. \$\endgroup\$ Aug 7 '20 at 15:52
-2
\$\begingroup\$

I would change this

for i in range(len(partitions)):
    items = list(partitions[i])   # convert from set
    for j in range(len(items)):
        edge_color_list[items[j]] = get_color((i+1)/len(partitions)

to:

for idx, partition in enumerate(partitions): # I personally do it your way 
    items = list(partition)   # convert from set
    for item in items: 
        edge_color_list[items] = get_color((idx+1)/len(partitions) 
        # idx is the `i` from before 

I referenced this.

There might be a linting plugin to help align your code towards being pythonic

\$\endgroup\$
3
  • \$\begingroup\$ feel free to edit/expand. I'm no expert... \$\endgroup\$ Aug 7 '20 at 16:29
  • \$\begingroup\$ This results in an error. Additionally how is this better than the OP's code? \$\endgroup\$
    – Peilonrayz
    Aug 7 '20 at 16:29
  • \$\begingroup\$ "feel free to edit/expand". The original title was "Writing python without a C accent" \$\endgroup\$ Aug 8 '20 at 21:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.