def test_X_Y_indep_Y_causes_MI_X(df_X_Y_indep_Y_causes_MI_X):
df = df_X_Y_indep_Y_causes_MI_X()
graph = MVPCStar(data=df).predict()
assert graph.get_bidirectional_edges() == set({})
assert graph.get_undirected_edges() == set({})
assert graph.get_marked_arrows() == set({('y', 'MI_x')})
assert graph.get_unmarked_arrows() == set({})
assert graph.get_nodes()
def test_Z_causes_X_Y_and_X_Z_causes_MI_Y(df_Z_causes_X_Y_and_X_Z_causes_MI_Y):
df = df_Z_causes_X_Y_and_X_Z_causes_MI_Y(size=10000)
graph = MVPCStar(data=df).predict()
assert graph.get_bidirectional_edges() == set({})
assert graph.get_undirected_edges() == set(
{frozenset({'x', 'z'}), frozenset({'z', 'y'})})
assert graph.get_marked_arrows() == set({('x', 'MI_y'), ('z', 'MI_y')})
assert graph.get_unmarked_arrows() == set({})
def test_long_chains_and_collider_with_MI(df_long_chains_and_collider_with_MI):
df = df_long_chains_and_collider_with_MI(size=50000)
graph = MVPCStar(data=df).predict()
assert graph.get_bidirectional_edges() == set({})
assert graph.get_undirected_edges() == set(
{frozenset({'a', 'b'}), frozenset({'e', 'd'})})
assert graph.get_unmarked_arrows() == set({('b', 'c'), ('d', 'c')})
assert graph.get_marked_arrows() == set({('c', 'MI_b')})
def test_chain_and_collider_without_MI(df_chain_and_collider_without_MI):
df = df_chain_and_collider_without_MI()
graph = MVPCStar(data=df).predict()
assert graph.get_bidirectional_edges() == set({})
assert graph.get_undirected_edges() == set(
{frozenset({'a', 'b'}), frozenset({'b', 'c'})})
assert graph.get_unmarked_arrows() == set({('a', 'd'), ('c', 'd')})
assert graph.get_marked_arrows() == set({})