Quick Start Guide

This guide will help you get started with MMSBM quickly.

Basic Usage

  1. First, import the necessary modules:

    import pandas as pd
from mmsbm import MMSBM

  2. Prepare your data:

    # Example data format
data = pd.DataFrame({
    'users': ['user1', 'user1', 'user2', 'user2'],
    'items': ['item1', 'item2', 'item1', 'item3'],
    'ratings': [5, 3, 4, 1]
})

  3. Initialize the model:

    model = MMSBM(
    user_groups=2,     # Number of user groups
    item_groups=4,     # Number of item groups
    iterations=500,    # Number of EM iterations
    sampling=5,        # Number of parallel runs
    seed=1            # For reproducibility
)

  4. Fit the model:

    model.fit(data)

  5. Make predictions:

    predictions = model.predict(test_data)

Complete Example

Here’s a complete example with synthetic data:

import pandas as pd
import numpy as np
from mmsbm import MMSBM

# Generate synthetic data
np.random.seed(42)
n_users = 100
n_items = 50

train_data = pd.DataFrame({
    'users': [f'user{i}' for i in np.random.randint(0, n_users, 1000)],
    'items': [f'item{i}' for i in np.random.randint(0, n_items, 1000)],
    'ratings': np.random.randint(1, 6, 1000)
})

# Initialize and fit model
model = MMSBM(user_groups=3, item_groups=5)
model.fit(train_data)

# Get model results
results = model.score()
print(f"Accuracy: {results['stats']['accuracy']:.3f}")
print(f"MAE: {results['stats']['mae']:.3f}")

Cross-Validation

To use cross-validation:

accuracies = model.cv_fit(data, folds=5)
print(f"Mean accuracy: {np.mean(accuracies):.3f} ± {np.std(accuracies):.3f}")

Model Parameters

Key parameters when initializing MMSBM:

  • user_groups: Number of user groups (K)

  • item_groups: Number of item groups (L)

  • iterations: Number of EM iterations

  • sampling: Number of parallel runs

  • seed: Random seed for reproducibility

  • debug: Enable debug logging

Additional Features

Parallel Processing

MMSBM automatically uses parallel processing for different sampling runs:

# Will use 10 parallel processes
model = MMSBM(sampling=10)
model.fit(data)

Debug Mode

For debugging and monitoring:

model = MMSBM(debug=True)
model.fit(data)

This will print additional information during fitting.

Next Steps