ysights.algorithms

YSights Algorithms

This module provides analysis algorithms for YSocial simulation data. It includes functions for analyzing agent profiles, detecting social paradoxes, evaluating recommendation systems, and studying topic dynamics.

Submodules:

• profiles: Agent profile and interest similarity analysis

• paradox: Visibility paradox detection and analysis

• recommenders: Recommendation system metrics and evaluation

• topics: Topic spread and adoption analysis

Example

Using algorithms to analyze simulation data:

from ysights import YDataHandler
from ysights.algorithms import (
    profile_topics_similarity,
    visibility_paradox,
    engagement_momentum
)

# Initialize data handler
ydh = YDataHandler('path/to/database.db')
network = ydh.social_network()

# Analyze profile similarity
similarities = profile_topics_similarity(ydh, network)

# Detect visibility paradox
paradox = visibility_paradox(ydh, network, N=100)
print(f"Paradox detected: {paradox['p_value'] < 0.05}")

# Calculate engagement momentum
momentum = engagement_momentum(ydh, time_window_rounds=24)

See also

• ysights.models: Data models and database interface

• ysights.viz: Visualization functions

Modules

paradox	Visibility Paradox Analysis
profiles	Agent Profile Analysis
recommenders	Recommender System Metrics
topics	Topic Analysis Algorithms