def model(base, dtc_data_vd):
base.cursor.execute("DROP MODEL IF EXISTS decision_tc_model_test")
base.cursor.execute(
"SELECT rf_classifier('decision_tc_model_test', 'public.dtc_data', 'TransPortation', '*' USING PARAMETERS exclude_columns='id, TransPortation', mtry=4, ntree=1, max_breadth=100, sampling_size=1, max_depth=6, nbins=40, seed=1, id_column='id')"
)
# I could use load_model but it is buggy
model_class = DecisionTreeClassifier(
"decision_tc_model_test",
cursor=base.cursor,
max_features=4,
max_leaf_nodes=100,
max_depth=6,
min_samples_leaf=1,
min_info_gain=0,
nbins=40,
)
model_class.input_relation = "public.dtc_data"
model_class.test_relation = model_class.input_relation
model_class.X = ["Gender", '"owned cars"', "cost", "income"]
model_class.y = "TransPortation"
base.cursor.execute(
"SELECT DISTINCT {} FROM {} WHERE {} IS NOT NULL ORDER BY 1".format(
model_class.y, model_class.input_relation, model_class.y))
classes = base.cursor.fetchall()
model_class.classes_ = [item[0] for item in classes]
yield model_class
model_class.drop()
def test_to_sql(self, model, titanic_vd):
model_test = DecisionTreeClassifier("rfc_sql_test")
model_test.drop()
model_test.fit(titanic_vd, ["age", "fare", "sex"], "survived")
current_cursor().execute(
"SELECT PREDICT_RF_CLASSIFIER(* USING PARAMETERS model_name = 'rfc_sql_test', match_by_pos=True)::int, {}::int FROM (SELECT 30.0 AS age, 45.0 AS fare, 'male' AS sex) x"
.format(model_test.to_sql()))
prediction = current_cursor().fetchone()
assert prediction[0] == prediction[1]
model_test.drop()
def test_contour(self, base, titanic_vd):
model_test = DecisionTreeClassifier("model_contour", cursor=base.cursor)
model_test.drop()
model_test.fit(
titanic_vd,
["age", "fare",],
"survived",
)
result = model_test.contour()
assert len(result.get_default_bbox_extra_artists()) == 34
model_test.drop()
def test_to_sql(self, model, titanic_vd):
model_test = DecisionTreeClassifier("rfc_sql_test", cursor=model.cursor)
model_test.drop()
model_test.fit(titanic_vd, ["age", "fare", "sex"], "survived")
model.cursor.execute(
"SELECT PREDICT_RF_CLASSIFIER(* USING PARAMETERS model_name = 'rfc_sql_test', match_by_pos=True, class=1, type='probability')::float, {}::float FROM (SELECT 30.0 AS age, 45.0 AS fare, 'male' AS sex) x".format(
model_test.to_sql()
)
)
prediction = model.cursor.fetchone()
assert prediction[0] == pytest.approx(prediction[1], 1e-2)
model_test.drop()
def test_to_python(self, model, titanic_vd):
model_test = DecisionTreeClassifier("rfc_python_test", cursor=model.cursor)
model_test.drop()
model_test.fit(titanic_vd, ["age", "fare", "sex"], "embarked")
model_test.cursor.execute(
"SELECT PREDICT_RF_CLASSIFIER(30.0, 45.0, 'male' USING PARAMETERS model_name = 'rfc_python_test', match_by_pos=True)"
)
prediction = model_test.cursor.fetchone()[0]
assert prediction == model_test.to_python(return_str=False)([[30.0, 45.0, 'male']])[0]
model_test.cursor.execute(
"SELECT PREDICT_RF_CLASSIFIER(30.0, 145.0, 'female' USING PARAMETERS model_name = 'rfc_python_test', match_by_pos=True)"
)
prediction = model_test.cursor.fetchone()[0]
assert prediction == model_test.to_python(return_str=False)([[30.0, 145.0, 'female']])[0]
def test_model_from_vDF(self, base, dtc_data_vd):
base.cursor.execute("DROP MODEL IF EXISTS tc_from_vDF")
model_test = DecisionTreeClassifier("tc_from_vDF", cursor=base.cursor)
model_test.fit(dtc_data_vd,
["Gender", '"owned cars"', "cost", "income"],
"TransPortation")
base.cursor.execute(
"SELECT model_name FROM models WHERE model_name = 'tc_from_vDF'")
assert base.cursor.fetchone()[0] == "tc_from_vDF"
model_test.drop()
def test_drop(self, base):
base.cursor.execute("DROP MODEL IF EXISTS decision_tc_model_test_drop")
model_test = DecisionTreeClassifier("decision_tc_model_test_drop",
cursor=base.cursor)
model_test.fit(
"public.dtc_data",
["Gender", '"owned cars"', "cost", "income"],
"TransPortation",
)
base.cursor.execute(
"SELECT model_name FROM models WHERE model_name = 'decision_tc_model_test_drop'"
)
assert base.cursor.fetchone()[0] == "decision_tc_model_test_drop"
model_test.drop()
base.cursor.execute(
"SELECT model_name FROM models WHERE model_name = 'decision_tc_model_test_drop'"
)
assert base.cursor.fetchone() is None
def test_repr(self, model):
assert "SELECT rf_classifier('public.decision_tc_model_test'," in model.__repr__()
model_repr = DecisionTreeClassifier("RF_repr")
model_repr.drop()
assert model_repr.__repr__() == "<RandomForestClassifier>"
