README

R framework for evaluating causal inference algorithms and for 
validating their estimates with a stability score ("stability 
score for causal inference", SSCI).

content:

test_utils.R 		functions for generation and simulation of SEMs, for comparing and handling graphs and results from differen algorithms for causal inference
scoring.R		functions for subsampling and calculation of the SSCI
example_code.R		some examples

violted_assumptions	experiments with violated assumptions, organised bottom up the files in the subfolders specify the sem and then call the upper files (sid_lin_gauss.R sid_exp_noise.R, sid_non_lin.R, sid_logistic.R, sid_unif_noise.R) to simulate the data and calculate the SID values with sid_samplesize.R. The effect interactions are specified in a seperate file for every sem in its own folder in the file interactions.R

SSCI_evaluation		
	score_propperties	
	evaluation
		all_seq_ssci.R				evaluation without interventions
		all_seq_intervention_ssci.R		evaluation with interventions
		plot_from_txt.R				functions for plotting the results of the evaluations
	
	score_properties
		compare_measures.R	comparison of different graph measueres based on random graphs
		ssci_random.R		experimental assesement of the properties of the SSCI with random graphs 
	

Author: Manuel Ruch