uv pip install l0-python

from l0 import SampleGate

# Select 1,000 households from 10,000
gate = SampleGate(n_samples=10000, target_samples=1000)
selected_data, indices = gate.select_samples(data)

# Gates learn which samples are most informative

from l0 import HardConcrete
from microcalibrate import Calibration

# L0 gates for household selection
gates = HardConcrete(
    len(household_weights),
    temperature=0.25,
    init_mean=0.999  # Start with most households
)

# Use in calibration
# microcalibrate applies gates during weight optimization

git clone https://github.com/PolicyEngine/L0
cd L0

tree l0/

# Key modules:
ls l0/
# - hard_concrete.py - Core L0 distribution
# - layers.py - L0Linear, L0Conv2d
# - gates.py - Sample/feature gates
# - penalties.py - L0/L2 penalty computation
# - temperature.py - Temperature scheduling

# Hard Concrete distribution (core algorithm)
cat l0/hard_concrete.py

# Sample gates (used in calibration)
cat l0/gates.py

# Neural network layers
cat l0/layers.py

cat l0/hard_concrete.py

from l0 import HardConcrete

# Create gates for household selection
gates = HardConcrete(
    n_items=len(households),
    temperature=0.25,
    init_mean=0.999  # Start with almost all households
)

# Gates produce probabilities (0 to 1)
probs = gates()

# Apply to weights during calibration
masked_weights = weights * probs

# See usage in data pipeline
grep -r "from l0 import" ../policyengine-us-data/

from l0 import TemperatureScheduler, update_temperatures

scheduler = TemperatureScheduler(
    initial_temp=1.0,  # Start relaxed
    final_temp=0.1,    # End sparse
    total_epochs=100
)

for epoch in range(100):
    temp = scheduler.get_temperature(epoch)
    update_temperatures(model, temp)
    # ... training ...

cat l0/temperature.py

from l0 import compute_l0l2_penalty

# Combine L0 (sparsity) with L2 (regularization)
penalty = compute_l0l2_penalty(
    model,
    l0_lambda=1e-3,  # Sparsity strength
    l2_lambda=1e-4   # Weight regularization
)

loss = task_loss + penalty

make test

# Or
pytest tests/ -v --cov=l0

cat tests/test_hard_concrete.py
cat tests/test_gates.py

from l0 import HybridGate

# Combine L0 selection with random sampling
hybrid = HybridGate(
    n_items=10000,
    l0_fraction=0.25,      # 25% from L0
    random_fraction=0.75,  # 75% random
    target_items=1000
)

selected, indices, types = hybrid.select(data)

from l0 import FeatureGate

# Select top features
gate = FeatureGate(n_features=1000, max_features=50)
selected_data, feature_indices = gate.select_features(data)

# Get feature importance
importance = gate.get_feature_importance()