def test_sklearn_tree1(self):
from sklearn.tree import DecisionTreeRegressor
X, y = iris_data()
check_model_representation(DecisionTreeRegressor(max_depth=1),
X,
y,
verbose=False,
suffix="t1")
def test_sklearn_tree2(self):
from sklearn.tree import DecisionTreeRegressor
X, y = iris_data()
check_model_representation(model=DecisionTreeRegressor(max_depth=2),
X=X,
y=y,
verbose=True,
suffix="t2")
def test_sklearn_scaler(self):
from sklearn.preprocessing import StandardScaler
data = numpy.array([[0, 0], [0, 0], [1, 1], [1, 1]],
dtype=numpy.float32)
check_model_representation(StandardScaler, data, verbose=False)
# The second compilation fails if suffix is not specified.
check_model_representation(model=StandardScaler,
X=data,
verbose=False,
suffix="_2")
def test_sklearn_linear_regression_verbose(self):
from sklearn.linear_model import LinearRegression
X, y = iris_data()
rows = []
def myprint(*args, **kwargs):
rows.append(' '.join(map(str, args)))
check_model_representation(
LinearRegression, X, y, verbose=True, fLOG=myprint)
self.assertGreater(len(rows), 2)
def test_sklearn_linear_regression_verbose(self):
X, y = iris_data()
rows = []
def myprint(*args, **kwargs):
rows.append(' '.join(map(str, args)))
check_model_representation(LinearRegression,
X,
y,
verbose=True,
fLOG=myprint,
suffix='A')
check_model_representation(LinearRegression,
X.tolist(),
y.tolist(),
verbose=True,
fLOG=myprint,
suffix='B')
self.assertGreater(len(rows), 2)
xdf = pandas.DataFrame(X)
try:
check_model_representation(LinearRegression,
xdf,
y.tolist(),
verbose=True,
fLOG=myprint,
suffix='B')
except TypeError as e:
self.assertIn("value is not a numpy.array but", str(e))
return
self.assertGreater(len(rows), 2)