files_plot.py

import json
import matplotlib.pyplot as plt
import networkx as nx
from collections import defaultdict
from matplotlib.lines import Line2D

# Load the data
with open('topic_rails.json', 'r') as file:
    data = json.load(file)

# Initialize a graph
G = nx.Graph()

# Initialize a dictionary to store related files per user
files_per_user = defaultdict(set)

# Process the data
for topic in data:
    entities = topic['entities_in_topic']
    for collaborator in topic['collaborators']:
        # Extract the user name without the role
        user = collaborator.split(' (')[0]
        files_per_user[user].update(entities)
        for entity in entities:
            G.add_edge(user, entity)

# Create labels with the number of related files
labels = {user: f"{user} ({len(files)})" for user, files in files_per_user.items()}
labels.update({file: file for file in G.nodes if file not in labels})

# Assign colors to nodes
node_colors = []
for node in G.nodes:
    if node in files_per_user:
        node_colors.append('skyblue')  # User nodes
    else:
        node_colors.append('lightgreen')  # File nodes

# Create a bipartite graph layout
users = list(files_per_user.keys())
files = [node for node in G.nodes if node not in users]
pos = nx.bipartite_layout(G, users)

# Plotting
plt.figure(figsize=(12, 8))
nx.draw(G, pos, labels=labels, with_labels=True, node_size=3000, node_color=node_colors, font_size=10, font_color='black', font_weight='bold', edge_color='gray')

# Create legend with circle shapes and bigger size
user_patch = Line2D([0], [0], marker='o', color='w', label='User', markersize=15, markerfacecolor='skyblue')
file_patch = Line2D([0], [0], marker='o', color='w', label='File', markersize=15, markerfacecolor='lightgreen')
plt.legend(handles=[user_patch, file_patch], loc='best')

plt.title('Related Files per User')
plt.savefig('related_files_per_user_graph.png')
plt.show()