mschauer / CausalInference.jl

Causal inference, graphical models and structure learning in Julia
https://mschauer.github.io/CausalInference.jl/latest/
Other
190 stars 24 forks source link

Implement sufficient adjustment criterion extending the backdoor criterion #60

Closed mschauer closed 1 year ago

mschauer commented 3 years ago

For DAGs: https://arxiv.org/pdf/1203.3515.pdf More general for CPDAG/MAG/PAG: https://auai.org/uai2015/proceedings/papers/155.pdf

mschauer commented 1 year ago

Is this for DAGs somehow included in #82 , @mwien?

mwien commented 1 year ago

Yes, find_covariate_adjustment finds a set $Z$ (with additional constraints $I$ and $R$, i.e. $I \subseteq Z \subseteq R$) which satisfies the adjustment criterion given in the Shpitser et al. paper (Def. 5) if such a set exists: https://github.com/mwien/CausalInference.jl/blob/ab5f4c9044045ebeaeeb5b4558d5cefc9dbe8b63/src/gensearch.jl#L159C1-L167

Similarly, find_min_covariate_adjustment and list_covariate_adjustment finds an inclusion minimal set satisfying this criterion resp. lists all such sets.

This is in contrast to function find_backdoor_adjustment, which find sets which satisfy the classical backdoor criterion by Pearl.

mschauer commented 1 year ago

Closed by #82