Replacing Sublists with Different Probabilities in GHPython

Grasshopper
,
1

I’m currently working on a GHPython script where I’ve successfully divided a list into sublists, each containing four elements. Now, I’m facing the challenge of replacing some of these sublists with other lists randomly, using different probabilities. For instance, I aim to replace the sublist [1, 0, 1, 1] with [1, 1, 1, 1] with a probability of 60%, [1, 2, 1, 1] with a probability of 20%, and [1, 3, 1, 1] with a probability of 20%.

Since GHPython doesn’t offer the random.choices() function and random.choice() doesn’t support specifying weights for selection, I’m seeking guidance on implementing this functionality without relying on external libraries like NumPy. Additionally, once I’ve successfully replaced these sublists, how can I structure the output as a data tree within Grasshopper, or at least ensure that the output doesn’t appear as a list of IronPython.Runtime.List objects?@Mahdiyar

2

You can make your own ‘choices’ function like this:

import random

def choices(population, weights, k=1):
    total_weight = float(sum(weights))
    cum_weights = [sum(weights[:i+1]) / total_weight for i in range(len(weights))]

    samples = []
    for _ in range(k):
        rand = random.random()
        for i, cum_weight in enumerate(cum_weights):
            if rand < cum_weight:
                samples.append(population[i])
                break

    return samples

# Example
a = choices(['A', 'B', 'C', 'D'], [0.1, 0.3, 0.2, 0.4], 10000)

Mahsa.gh (10.9 KB)

Mahsa
Mahsa4285×1215 329 KB

import ghpythonlib.treehelpers as th
my_list = [[1, 2, 3], [3, 4]]
a = th.list_to_tree(my_list)
2 Likes
3

Thanks a lot!