Tutorial 2: Data Generation

Learn how to generate synthetic causal datasets for training and evaluation.

Coming Soon

This tutorial is under development. For now, see:

Data Generation - Complete data generation guide
Tutorial 1: Basics - Basic concepts
Standard CATE Estimation Example - Complete example

Quick Reference

Standard CATE Data

from causalfm.data import StandardCATEGenerator

generator = StandardCATEGenerator(
    num_samples=1024,
    num_features=10,
    seed=42
)

# Single dataset
df = generator.generate()

# Multiple datasets
generator.generate_multiple(100, "data/train/")

IV Data

from causalfm.data import IVDataGenerator

generator = IVDataGenerator(
    num_samples=1024,
    num_features=10,
    instrument_type='binary',
    seed=42
)

df = generator.generate()

Front-door Data

from causalfm.data import FrontdoorDataGenerator

generator = FrontdoorDataGenerator(
    num_samples=1024,
    num_features=10,
    num_confounders=5,
    seed=42
)

df = generator.generate()

Next Tutorial

Continue to Tutorial 3: Training Models to learn about model training.