Getting Started

• Getting started with DoWhy: A simple example
  • Interface 1 (recommended): Input causal graph
  • Interface 2: Specify common causes and instruments
  • Refuting the estimate
• Confounding Example: Finding causal effects from observed data
  • Let’s create a mystery dataset for which we need to determine whether there is a causal effect.
  • Using DoWhy to resolve the mystery: Does Treatment cause Outcome?
• DoWhy: Different estimation methods for causal inference
  • Identifying the causal estimand
  • Method 1: Regression
  • Method 2: Distance Matching
  • Method 3: Propensity Score Stratification
  • Method 4: Propensity Score Matching
  • Method 5: Weighting
  • Method 6: Instrumental Variable
  • Method 7: Regression Discontinuity
• Simple example on using Instrumental Variables method for estimation
  • Loading the dataset
  • Using DoWhy to estimate the causal effect of education on future income
• Different ways to load an input graph
  • I. Generating dummy data
  • II. Loading GML or DOT graphs
• DoWhy: Interpreters for Causal Estimators
  • Identifying the causal estimand
  • Method 1: Propensity Score Stratification
  • Method 2: Propensity Score Matching
  • Method 3: Weighting
• Demo for the DoWhy causal API
  • Comparing the estimate to Linear Regression
• Do-sampler Introduction
  • Pearlian Interventions
  • Statefulness
  • Integration
  • Specifying Interventions
  • Demo