Algorithms

Apart from the original parameters of the underlying software, each algorithm module is equipped with an additional parameter, timeout, which is the maximum time in seconds allowed for the algorithm to run. After the timeout, the algorithm will be terminated and either an empty file will be created or the current best graph will be saved (if the algorithm supports that).

Modules for MCMC algorithms can be used seamlessly with the other modules. However, apart from the original parameters and timeout, these modules have four additional fields:

• mcmc_seed is the random seed for the algorithm.

• mcmc_estimator specifies which estimator to use (threshold or map).

• threshold specifies the threshold for the posterior edge probabilities if mcmc_estimator is set to threshold.

• burnin_frac is a value in (0, 1) that specifies the fraction of the samples at the beginning of the graph trajectory to be discarded as burn-in.

The available modules are listed below. To add new modules, see Adding new modules.

Algorithm	Graph	Package	Module
Chordal graph samplers	DG	Alun Thomas	athomas_jtsamplers
BDgraph	UG	BDgraph	bdgraph
Iterative MCMC	DAG, CPDAG	BiDAG	bidag_itsearch
Order MCMC	DAG, CPDAG	BiDAG	bidag_order_mcmc
Partition MCMC	DAG, CPDAG	BiDAG	bidag_partition_mcmc
Fast IAMB	DAG	bnlearn	bnlearn_fastiamb
Grow-shrink	DAG	bnlearn	bnlearn_gs
H2PC	DAG	bnlearn	bnlearn_h2pc
HC	DAG	bnlearn	bnlearn_hc
HPC	DAG	bnlearn	bnlearn_hpc
IAMB	DAG	bnlearn	bnlearn_iamb
IAMB-FDR	DAG	bnlearn	bnlearn_iambfdr
INTER-IAMB	DAG	bnlearn	bnlearn_interiamb
MMHC	DAG	bnlearn	bnlearn_mmhc
MMPC	DAG	bnlearn	bnlearn_mmpc
PC	DAG	bnlearn	bnlearn_pcstable
RSMAX2	DAG	bnlearn	bnlearn_rsmax2
S-I HITON-PC	DAG	bnlearn	bnlearn_sihitonpc
Tabu	DAG	bnlearn	bnlearn_tabu
GSP	DAG	CausalDAG	causaldag_gsp
GRaSP	CPDAG	causal-learn	causallearn_grasp
Corrmat thresh	UG	Benchpress	corr_thresh
Dual PC	CG, CPDAG	dualPC	dualpc
Psi-learning	UG	equSA	equsa_psilearner
ANM	DAG	gCastle	gcastle_anm
CORL	DAG	gCastle	gcastle_corl
Direct LINGAM	DAG	gCastle	gcastle_direct_lingam
GAE	DAG	gCastle	gcastle_gae
GOLEM	DAG	gCastle	gcastle_golem
GraNDAG	DAG	gCastle	gcastle_grandag
ICALiNGAM	DAG	gCastle	gcastle_ica_lingam
MCSL	DAG	gCastle	gcastle_mcsl
NO TEARS	DAG	gCastle	gcastle_notears
NO TEARS low rank	DAG	gCastle	gcastle_notears_low_rank
NO TEARS non-linear	DAG	gCastle	gcastle_notears_nonlinear
PC	DAG	gCastle	gcastle_pc
RL	DAG	gCastle	gcastle_rl
GOBNILP	DAG	GOBNILP (BitBucket)	gobnilp
GrUES	UDG	gues	grues
Graphical lasso	UG	huge	huge_glasso
Meinshausen & Buhlmann cov est	UG	huge	huge_mb
TIGER	UG	huge	huge_tiger
Parallel DG	DG	parallelDG	paralleldg
GIES	CPDAG	pcalg	pcalg_gies
PC	CPDAG, CG	pcalg	pcalg_pc
Precmat thresh	UG	Benchpress	prec_thresh
ASOBS	DAG	r.blip	rblip_asobs
Graphical Lasso	UG	scikit-learn	sklearn_glasso
BOSS	CPDAG	causal-cmd	tetrad_boss
FASK	DAG	causal-cmd	tetrad_fask
FGES	CPDAG	causal-cmd	tetrad_fges
FOFC	DAG	causal-cmd	tetrad_fofc
FTFC	DAG	causal-cmd	tetrad_ftfc
GRaSP	CPDAG	causal-cmd	tetrad_grasp
ICA-LINGAM	DAG	causal-cmd	tetrad_ica-lingam
PC	DAG	causal-cmd	tetrad_pc
Particle Gibbs	DG	trilearn	trilearn_pgibbs