Exogenous variable
An exogenous variable is a causal input whose variation originates outside the endogenous system represented by a structural causal model.
Causal AI
An exogenous cause is a causal input whose variation originates outside the endogenous system represented by a structural causal model.
- How it is used
- Represents unmodeled causal inputs when constructing generative structural causal models and inferring counterfactual outcomes.
- How deep it goes
- Defines latent root inputs, works through distributions and deterministic assignments, and qualifies independence and identifiability assumptions.
- Examples
- Uses a Pyro femur-height SCM, a rule-based rock-throwing process, and random inputs to a door-choice game.
- What you can do after
- Build it from scratch by specifying exogenous distributions and assignments, then infer latent inputs for counterfactual prediction.
Causal Inference for Data Science
An exogenous variable is a variable whose variation is determined outside the modeled causal system and is not explained by its included predictors.
- How it is used
- Marks the boundary between modeled causal variables and unobserved influences from outside the system.
- How deep it goes
- Defines exogenous inputs, embeds them in structural functions, and applies conditional-mean assumptions without identifying their exact laws.
- Examples
- Uses UZ, UX, UY, and the price–sales variables D and O to illustrate external influences in causal equations.
- What you can do after
- Evaluate/understand causal models by separating modeled relationships from unobserved variation and stating assumptions about that variation.
Read how Causal Inference for Data Science explains exogenous variable →
Related concepts
- structural causal modelpart of · A structural causal model represents causal relationships among variables through equations that determine each variable from other variables and exogenous factors.
- canonical distributionconfigured by · A canonical distribution is a standard mathematically characterized probability distribution used as a primitive for probabilistic models.
- parallel world graphused by · A parallel world graph is a graph representation that places variables from multiple counterfactual worlds together without necessarily permitting the same separation relations as a counterfactual graph.
- abductionused by · Abduction is the inference step that uses observed facts to infer the latent circumstances or exogenous values that could have produced them.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.