skills.rest
xjtulyc
xjtulycCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

causal-inference

Estimate causal treatment effects from observational datasets using regression, propensity score weighting, and doubly robust methods.

33|6|Updated Mar 17, 2026
Search Tags:#causal inference#causal dag#propensity score#ate estimation#doubly robust#refutation tests
One-click install
npx skills add https://github.com/xjtulyc/awesome-rosetta-skills --skill causal-inference-xjtulyc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: causal-inference
Source: https://github.com/xjtulyc/awesome-rosetta-skills/tree/main/skills/03-mathematics/causal-inference
Command: npx skills add https://github.com/xjtulyc/awesome-rosetta-skills --skill causal-inference-xjtulyc

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires dowhy, causalml, networkx, scikit-learn, pandas, numpy, matplotlib, statsmodels.

What problem does it solve?

This Skill helps you turn observational data into defensible estimates of causal effects rather than misleading correlations.

Core Features & Use Cases

  • Causal DAG design: Build and visualize causal graphs to reason about confounding and identification assumptions.
  • Effect identification & estimation: Estimate ATE/ATT/CATE-like quantities using multiple approaches (regression, IPW, doubly robust) consistent with identification via adjustment sets.
  • Validation via refutation: Stress-test estimated effects using placebo tests, common-cause perturbations, and subset re-estimation.
  • Use Case: You have a dataset where a “treatment” (e.g., policy adoption or education level) was not randomized, and you want to estimate the treatment’s impact on an outcome with careful adjustment and diagnostics.

Quick Start

Use causal-inference to estimate the ATE (and confidence interval) for a treatment column against an outcome column while adjusting for specific confounders.

Frequently Asked Questions about causal-inference

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I estimate causal treatment effects from observational data?▼

To estimate causal treatment effects from observational data, you can use causal DAG reasoning alongside regression, propensity score weighting, and doubly robust methods to calculate ATE or ATT while adjusting for specified confounders.

What is the best way to validate ATE estimation results against hidden confounders?▼

Validating ATE estimation results involves running refutation tests like placebo treatments, common-cause perturbations, and subset re-estimation to stress-test the robustness of your estimated causal effects against hidden confounders.

Can I use propensity score weighting and doubly robust estimation for policy evaluation?▼

Yes, propensity score weighting and doubly robust estimation are supported for policy evaluation, clinical effectiveness research, and economics studies to estimate treatment impacts under confounding and overlap assumptions.

How do I build a causal DAG to identify confounders for my dataset?▼

Building a causal DAG allows you to reason about confounding and identification assumptions, visually mapping relationships to determine the correct adjustment sets required for estimating causal effects.

Does this approach require randomized treatment assignment to work?▼

No, it does not require randomized treatment assignment; it is specifically designed to estimate causal effects from observational datasets where treatments were not randomized, using backdoor-style adjustment logic.

What libraries are needed for causal inference and refutation testing workflows?▼

Causal inference and refutation testing workflows utilize libraries including dowhy, causalml, networkx, scikit-learn, pandas, numpy, matplotlib, and statsmodels for graph reasoning and effect estimation.

Related Skills

Explore More Education & Research Skills→
rocklambrosrocklambros
2

analyzing-causal-dag

Determine valid adjustment sets from observational causal DAGs using the backdoor criterion.

Community
Advanced
xjtulycxjtulyc
33

ml-econometrics

Estimate causal treatment effects with Post-LASSO and DML cross-fitting.

Community
Advanced
eyadsibaieyadsibai
7

data-science

Provide Python code for statistical modeling, A/B testing, and feature engineering.

Community
Advanced
tondevreltondevrel
19

dowhy

Estimate causal effects from observational data using the identify-estimate-refute framework.

Community
Advanced