Julia package for causal inference, graphical models and structure learning.
This package contains code for the stable version of the PC algorithm and the FCI algorithm as described in Zhang's article.
The algorithms use the Julia packagea LightGraphs and MetaGraphs. Graphs are represented by sorted adjacency lists (vectors in the implemention). CPDAGs are just
DiGraphs where unoriented edges are represented by both a forward and a backward directed edge. PAGs are
MetaDiGraphs where every edge is represented by a forward and a backward edge. The marks of the endpoint of an edge are stored in the
:marks property. Marks can be checked and set using the
set_marks! functions (see documentation for details).
The PC algorithm is tested on random DAGs by comparing the result of the PC algorithm using the d-separation oracle with CPDAGs computed with Chickering's DAG->CPDAG conversion algorithm (implemented as
cpdag in this package).
After examples discussed in chapter 2 of Pearl, Judea. Causality. Cambridge University Press, 2009.
pc.jl in the example directory.
using CausalInference using LightGraphs include("plotdag.jl") # Generate some data N = 1000 p = 0.01 x = randn(N) v = x + randn(N)*0.25 w = x + randn(N)*0.25 z = v + w + randn(N)*0.25 s = z + randn(N)*0.25 df = (x=x, v=v, w=w, z=z, s=s) println("Running Gaussian tests") @time est_g = pcalg(df, p, gausscitest) variables = [String(k) for k in keys(df)] tp = plot_dag(est_g, variables) save(PDF("estdag"), tp)
Not all causal directions are indentified (and identifiable) in this example, and visualized by edges with circled/unknown arrow marks.
But we can conclude without intervention from observations alone that for example
V are causal for
Q: I looked for "causal inference" and found CausalInference.jl and Omega.jl... A: CausalInference.jl is about causal discovery, you observe the data and want to infer the causal structure. Omega lets you reason what happens then: when you intervene ("do calculus") and want to cause changes.
- P. Spirtes, C. Glymour, R. Scheines, R. Tillman: Automated search for causal relations: Theory and practice. Heuristics, Probability and Causality: A Tribute to Judea Pearl 2010
- P. Spirtes, C. Glymour, R. Scheines: Causation, Prediction, and Search. MIT Press 2000
- J. Zhang: On the completeness of orientation rules for causal discovery in the presence of latent confounders and selection bias. Artificial Intelligence 16-17 (2008), 1873-1896
- T. Richardson, P. Spirtes: Ancestral Graph Markov Models. The Annals of Statistics 30 (2002), 962-1030
- D. M. Chickering: Learning Equivalence Classes of Bayesian-Network Structures. Journal of Machine Learning Research 2 (2002), 445-498.
- D. Colombo, M. H. Maathuis: Order-Independent Constraint-Based Causal Structure Learning. Journal of Machine Learning Research 15 (2014), 3921-3962.