econml
This package contains several methods for calculating Conditional Average Treatment Effects
Decision gist · record as of 2026-08-14
Yes, if you need to estimate causal treatment effects from observational data and want flexibility in modeling effect heterogeneity. The package is actively maintained, well-documented, MIT-licensed, and has no known vulnerabilities. Install friction is moderate due to dependencies; ensure your environment can build numba and scipy. Not suitable if you only need average treatment effects or lack domain knowledge of causal assumptions.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python ≥3.9; medium install friction due to compiled dependencies (numba, scipy, lightgbm) that may require build tools on some systems.
- Medium install friction due to 11 runtime dependencies including numpy, scipy, numba, and lightgbm.
- Active maintenance with recent release (14 days old) and strong repository signals (4753 stars, last commit 2026-08-11).
License · maintenance · safety
MIT (permissive) — MIT license permits commercial and private use with minimal restrictions; you may use, modify, and distribute the package freely provided you include the license notice.
last release 2026-07-31 (14 days) · last repo commit 2026-08-11 · 4,753 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 563,080 downloads/mo, #5,982 on PyPI
Alternatives
Verify before relying
pip install econml
from econml.dml import LinearDML
from econml.inference import BootstrapInference
est = LinearDML(model_y=LassoCV(), model_t=LassoCV())
est.fit(Y, T, X=X, W=W)- Whether the package supports GPU acceleration or distributed computation for large datasets.
- Performance characteristics and scalability limits for high-dimensional feature spaces.
- Specific causal assumptions required by each estimation method and how to validate them.
- Whether LassoCV is available as a built-in or requires external installation.
What it is and what it does
EconML is a machine learning toolkit for causal inference that estimates heterogeneous treatment effects—how the causal impact of an intervention on an outcome varies across different groups or feature values. It combines econometric methods with modern ML techniques to work with observational (non-experimental) data, where you measure treatment T's effect on outcome Y while controlling for confounders X and W. The package provides a unified API across multiple estimation strategies, many of which return confidence intervals and statistical inference results.
The core use case is personalized decision-making: given a sample with particular features, what is the expected causal effect of applying a treatment to that sample? This is useful in policy learning, A/B testing analysis, and any domain where you need to understand not just average effects but how effects differ across populations. The package assumes either no unobserved confounders or access to instrumental variables, depending on the method chosen.
Use it for
- Estimate how a marketing intervention's effect on customer spending varies by customer demographics or purchase history.
- Analyze clinical trial or observational medical data to determine which patient subgroups benefit most from a treatment.
- Evaluate policy interventions to identify which populations see the largest causal effects.
- Learn optimal treatment assignment policies by combining effect estimation with policy learning methods.
- Conduct causal model selection and cross-validation to choose among competing heterogeneous effect estimators.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you need to estimate causal treatment effects from observational data and want flexibility in modeling effect heterogeneity.
The package is actively maintained, well-documented, MIT-licensed, and has no known vulnerabilities. Install friction is moderate due to dependencies; ensure your environment can build numba and scipy. Not suitable if you only need average treatment effects or lack domain knowledge of causal assumptions.
Install
econml on PyPI
Before you install
Medium install friction due to 11 runtime dependencies including numpy, scipy, numba, and lightgbm. Active maintenance with recent release (14 days old) and strong repository signals (4753 stars, last commit 2026-08-11). Prebuilt wheels available for Python 3.9–3.14 across macOS, Linux, and Windows.
Requires Python ≥3.9; medium install friction due to compiled dependencies (numba, scipy, lightgbm) that may require build tools on some systems.
License in practice
MIT license permits commercial and private use with minimal restrictions; you may use, modify, and distribute the package freely provided you include the license notice.
Quickstart
pip install econml
from econml.dml import LinearDML
from econml.inference import BootstrapInference
est = LinearDML(model_y=LassoCV(), model_t=LassoCV())
est.fit(Y, T, X=X, W=W)Verify before relying
- Whether the package supports GPU acceleration or distributed computation for large datasets.
- Performance characteristics and scalability limits for high-dimensional feature spaces.
- Specific causal assumptions required by each estimation method and how to validate them.
- Whether LassoCV is available as a built-in or requires external installation.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.9 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 11 packagesnumpynumbascipyscikit-learnsparsejoblibstatsmodelspandasshaplightgbmpackaging |
| Maintenance | Actively maintained 14 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 563,080 / month, #5,982 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Operating System :: MacOSOperating System :: Microsoft :: WindowsOperating System :: POSIX :: LinuxProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Programming Language :: Python :: 3.9 |
Evidence: econml-0.17.0-cp310-cp310-macosx_11_0_arm64.whl; econml-0.17.0-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; econml-0.17.0-cp310-cp310-win_amd64.whl; econml-0.17.0-cp311-cp311-macosx_11_0_arm64.whl; econml-0.17.0-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; econml-0.17.0-cp311-cp311-win_amd64.whl; econml-0.17.0-cp312-cp312-macosx_11_0_arm64.whl; econml-0.17.0-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; econml-0.17.0-cp312-cp312-win_amd64.whl; econml-0.17.0-cp313-cp313-macosx_11_0_arm64.whl; econml-0.17.0-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; econml-0.17.0-cp313-cp313-win_amd64.whl; econml-0.17.0-cp314-cp314-macosx_11_0_arm64.whl; econml-0.17.0-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; econml-0.17.0-cp314-cp314t-macosx_11_0_arm64.whl; econml-0.17.0-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; econml-0.17.0-cp314-cp314t-win_amd64.whl; econml-0.17.0-cp314-cp314-win_amd64.whl; econml-0.17.0-cp39-cp39-macosx_11_0_arm64.whl; econml-0.17.0-cp39-cp39-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “heterogeneous treatment effects”
- econmlEconML estimates heterogeneous treatment effects from observational…
- causalmlCausalml estimates the causal impact of treatments on outcomes at the…
- causallibCausallib estimates causal effects from observational data using a…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also causalml · causallib · cem · pyfixest · psmpy · rdrobust · dowhy · pyhdfe · empirical-calibration · linearmodels