dodiscover.constraint.LearnSkeleton#
- class dodiscover.constraint.LearnSkeleton(ci_estimator, sep_set=None, alpha=0.05, min_cond_set_size=0, max_cond_set_size=None, max_combinations=None, condsel_method=ConditioningSetSelection.NBRS, keep_sorted=False, n_jobs=None)[source]#
Learn a skeleton graph from observational data without latent confounding.
A skeleton graph from a Markovian causal model can be learned completely with this procedure.
- Parameters:
- ci_estimator
BaseConditionalIndependenceTest The conditional independence test function.
- sep_setdictionary of dictionary of
listofset Mapping node to other nodes to separating sets of variables. If
None, then an empty dictionary of dictionary of list of sets will be initialized.- alpha
float, optional The significance level for the conditional independence test, by default 0.05.
- min_cond_set_size
int The minimum size of the conditioning set, by default 0. The number of variables used in the conditioning set.
- max_cond_set_size
int, optional Maximum size of the conditioning set, by default None. Used to limit the computation spent on the algorithm.
- max_combinations
int, optional The maximum number of conditional independence tests to run from the set of possible conditioning sets. By default None, which means the algorithm will check all possible conditioning sets. If
max_combinations=nis set, then for every conditioning set size, ‘p’, there will be at most ‘n’ CI tests run before the conditioning set size ‘p’ is incremented. For controlling the size of ‘p’, seemin_cond_set_sizeandmax_cond_set_size. This can be used in conjunction withkeep_sortedparameter to only test the “strongest” dependences.- condsel_method
ConditioningSetSelection The method to use for selecting the conditioning set. Must be one of (‘complete’, ‘neighbors’, ‘neighbors_path’). See Notes for more details.
- keep_sortedbool
Whether or not to keep the considered conditioning set variables in sorted dependency order. If True (default) will sort the existing dependencies of each variable by its dependencies from strongest to weakest (i.e. largest CI test statistic value to lowest). This can be used in conjunction with
max_combinationsparameter to only test the “strongest” dependences.- n_jobs
int, optional Number of CPUs to use, by default None.
- ci_estimator
Notes
Proceed by testing neighboring nodes, while keeping track of test statistic values (these are the ones that are the “most dependent”). Remember we are testing the null hypothesis
\[H_0: X \perp Y | Z\]where the alternative hypothesis is that they are dependent and hence require a causal edge linking the two variables.
Different methods for learning the skeleton:
There are different ways to learn the skeleton that are valid under various assumptions. The value of
condsel_methodcompletely defines how one selects the conditioning set.‘complete’: This exhaustively conditions on all combinations of variables in the graph. This essentially refers to the SGS algorithm [1]
‘neighbors’: This only conditions on adjacent variables to that of ‘x_var’ and ‘y_var’. This refers to the traditional PC algorithm [2]
‘neighbors_path’: This is ‘neighbors’, but restricts to variables with an adjacency path from ‘x_var’ to ‘y_var’. This is a variant from the RFCI paper [3]
- Attributes:
- adj_graph_
nx.Graph The discovered graph from data. Stored using an undirected graph. The graph contains edge attributes for the smallest value of the test statistic encountered (key name ‘test_stat’), the largest pvalue seen in testing ‘x’ || ‘y’ given some conditioning set (key name ‘pvalue’).
- sep_set_dictionary of dictionary of
listofset Mapping node to other nodes to separating sets of variables.
- context_
Context The result context. Encodes causal assumptions.
- min_cond_set_size_
int The inferred minimum conditioning set size.
- max_cond_set_size_
int The inferred maximum conditioning set size.
- max_combinations_
int The inferred maximum number of combinations of ‘Z’ to test per \(X \perp Y | Z\).
- n_iters_
int The number of iterations the skeleton has been learned.
- adj_graph_
Methods
evaluate_edge(data, conditional_test_func, X, Y)Test any specific edge for X || Y | Z.
learn_graph
- ci_estimator#
Callable[[Column, Column, Set[Column]], Tuple[float, float]]
- evaluate_edge(data, conditional_test_func, X, Y, Z=None)#
Test any specific edge for X || Y | Z.