ml4t-engineer

Feature engineering, labeling, alternative bars, and leakage-safe datasets for financial ML.

ml4t-engineer provides 120 registry features across 11 categories, path-dependent and fixed-horizon labeling, activity-based bar sampling, and train-only preprocessing. The core interface uses Polars DataFrames.

Installation

ml4t-engineer supports Python 3.12, 3.13, and 3.14 on Linux, macOS, and Windows.

pip install ml4t-engineer

The quick start uses only core dependencies. Optional extras provide TA-Lib validation, DuckDB and PyArrow storage, market calendars, visualization, statistics, and ML tools:

pip install "ml4t-engineer[ta]"
pip install "ml4t-engineer[store]"
pip install "ml4t-engineer[calendars]"
pip install "ml4t-engineer[viz]"
pip install "ml4t-engineer[stats]"
pip install "ml4t-engineer[ml]"

TA-Lib requires its native library. The core package does not require an external service, credentials, or special hardware. Python 3.15 is not supported while the active Polars compatibility exception applies.

Quick Start

from datetime import date, timedelta

import polars as pl
from ml4t.engineer import compute_features

close = [100.0 + i * 0.1 + (i % 7) * 0.2 for i in range(100)]
ohlcv = pl.DataFrame(
    {
        "timestamp": [date(2024, 1, 1) + timedelta(days=i) for i in range(100)],
        "open": close,
        "high": [price + 1.0 for price in close],
        "low": [price - 1.0 for price in close],
        "close": close,
        "volume": [100_000 + i * 100 for i in range(100)],
    }
)

features = compute_features(ohlcv, ["rsi", "macd", "atr"])

assert {"rsi", "macd", "atr"} <= set(features.columns)
assert features.height == ohlcv.height

compute_features() returns the input columns with the requested feature columns appended. Use the feature registry to inspect categories and parameters before building larger pipelines.

Supported Workflows

• Technical, volatility, risk, microstructure, statistical, and ML-oriented features
• Triple-barrier, ATR-barrier, percentile, fixed-horizon, trend-scanning, and meta-labels
• Tick, volume, dollar, imbalance, and run bars
• Train/test splitting with train-only scaling
• Feature metadata search and discovery

See the documentation for tutorials, task-oriented guides, explanations, and the API reference. Report defects and request changes through GitHub Issues.

Related Libraries

• ml4t-specs defines the shared market-data and artifact contracts used by this package.
• ml4t-data supplies validated market data for feature computation.
• ml4t-diagnostic evaluates features, labels, and model signals produced from engineered datasets.

Development

git clone https://github.com/ml4t/engineer.git
cd engineer
uv sync --dev --extra docs --extra ta --extra store --extra viz
uv run ruff check src/ tests/ examples/ scripts/
uv run ruff format --check src/ tests/ examples/ scripts/
uv run ty check
uv run pytest tests/ -q
uv build
uv run mkdocs build --strict

Pull requests must also pass the supported Python and operating-system matrix, dependency and vulnerability review, clean-wheel installation, documented workflow tests, and ecosystem qualification.

Project Information

• Documentation
• Issue tracker
• Releases and changelog
• License