SMAdiffz 1.1.1

Powerful data structures for data analysis, time series for information diffusion analysis

The proposed package is designed to facilitate comprehensive work on information diffusion analysis. It provides a versatile set of tools and functionalities that empower users to explore, model, and analyze the intricate dynamics of information spread within a given system.

Following is a sample scenario.

def measure_information_diffusion(posts, threshold): # Sort the set of posts . This should using timestamps posts = sorted(posts, key=lambda x: x.timestamp)

# Initialize empty set of trees
trees = set()

# Iterate over each post in the sorted set
for i, p_i in enumerate(posts):
    # Initialize a new tree with a single node representing (p_i)
    T_i = {p_i}

    # For each post with a timestamp later than p_i
    for j in range(i + 1, len(posts)):
        p_j = posts[j]

        # Compute the similarity between the tags of p_i and p_j
        similarity = compute_similarity(p_i, p_j)

        # If similarity is above the threshold, add a directed edge from p_i to p_j in T_i
        if similarity > threshold:
            T_i.add(p_j)

        # If p_j has already been added to a diffusion tree in trees, merge T_i with that tree
        for T_j in trees:
            if p_j in T_j:
                T_j.update(T_i)
                break
        else:
            # If p_j hasn't been added to any diffusion tree, add T_i to trees
            trees.add(T_i)

# Return the set of diffusion trees
return trees