def inference_dag(self):
"""
This is a DAG created from the original but has been altered
to accomodate `do-calculus`.
"""
infer_dag = DAG(self.dag.df.copy())
logging.debug(
f"constructing copy of original DAG nodes: {infer_dag.nodes}")
for n1, n2 in self.dag.edges:
if n2 not in self.do_dict.keys():
infer_dag.add_edge(n1, n2)
else:
logging.debug(
f"edge {n1} -> {n2} ignored because of do operator")
logging.debug(f"original DAG copied")
return infer_dag
def simple_dag():
df = pd.DataFrame({
"a": [1, 1, 1, 1, 0, 0, 0, 0],
"b": [0, 1, 0, 1, 1, 1, 1, 0],
"c": [0, 0, 1, 0, 0, 1, 0, 1]
})
return DAG(df).add_edge("a", "b").add_edge("a", "c").add_edge("c", "b")
def basic_dag():
df = pd.DataFrame({"a": [1, 1, 1, 1, 0, 0, 0, 0],
"b": [0, 1, 0, 1, 1, 1, 1, 0],
"c": [0, 0, 1, 0, 0, 1, 0, 1],
"d": [1, 1, 0, 1, 0, 0, 0, 0],
"e": [1, 1, 1, 1, 0, 0, 0, 0]})
return DAG(df)
def test_parent_child1(small_df):
dag = DAG(small_df).add_edge("a", "b").add_edge("c",
"b").add_edge("a", "c")
assert set(dag.children("a")) == {"b", "c"}
assert set(dag.children("b")) == set()
assert set(dag.children("c")) == {"b"}
assert set(dag.parents("a")) == set()
assert set(dag.parents("b")) == {"a", "c"}
assert set(dag.parents("c")) == {"a"}
def test_copy(small_df):
dag1 = DAG(small_df).add_edge("a", "b").add_edge("c", "b")
dag2 = DAG(small_df).add_edge("a", "b").add_edge("c",
"b").add_edge("a", "c")
dag1_copy = dag1.copy()
dag2_copy = dag2.copy()
assert set(dag1_copy.edges) == set(dag1.edges)
assert set(dag2_copy.edges) == set(dag2.edges)
assert set(dag1_copy.nodes) == set(dag1.nodes)
assert set(dag2_copy.nodes) == set(dag2.nodes)
assert dag1.df.equals(dag1_copy.df)
assert dag2.df.equals(dag2_copy.df)
import logging
import pandas as pd
from brent.graph import DAG
from brent.query import Query
logging.basicConfig(
level=logging.DEBUG,
format='%(asctime)s [%(filename)s:%(funcName)s:%(lineno)d] %(levelname)s - %(message)s',
)
df = pd.DataFrame({"a": [1, 1, 1, 1, 0, 0, 0, 0],
"b": [0, 1, 0, 1, 1, 1, 1, 0],
"c": [0, 0, 1, 0, 0, 1, 0, 1]})
dag = DAG(df).add_edge("a", "b").add_edge("a", "c").add_edge("c", "b")
logging.debug(f"parents of c: {dag.parents('c')}")
for node in dag.nodes:
logging.debug(f"confirming parents({node}) = {dag.parents(node)}")
print(Query(dag).infer())
def dag_with_two_values():
return (DAG(make_fake_df(nodes=4))
.add_edge("a", "d")
.add_edge("b", "d")
.add_edge("a", "b")
.add_edge("b", "c"))
def dag_with_four_values():
return (DAG(make_fake_df(nodes=4, rows=2000, values=4))
.add_edge("a", "d")
.add_edge("b", "d")
.add_edge("a", "b")
.add_edge("b", "c"))
def test_dag_remains_dag(small_df):
dag = DAG(small_df).add_edge("a", "b").add_edge("b", "c")
with pytest.raises(ValueError):
assert dag.add_edge("c", "a")
def test_connections2(small_df):
dag = DAG(small_df).add_edge("a", "b").add_edge("c", "b")
assert set(dag.connections("a")) == {"b"}
assert set(dag.connections("b")) == {"a", "c"}
assert set(dag.connections("c")) == {"b"}
def test_origin_nodes_2(small_df):
dag = DAG(small_df).add_edge("a", "b").add_edge("c", "b")
assert dag.origin_nodes == ("a", "c")
def dag():
return (DAG(make_fake_df(7)).add_edge("e", "a").add_edge(
"e",
"d").add_edge("a", "d").add_edge("b", "d").add_edge("a", "b").add_edge(
"a", "c").add_edge("b", "c").add_edge("c", "f").add_edge("g", "f"))