def test_classify_ada_multiclass(self):
print("TestBDTsklearn test_classify_ada_multiclass")
model = AdaBoostClassifier(DecisionTreeClassifier(max_depth=3),
n_estimators=5,
learning_rate=1)
model.fit(self.data_x, self.data_y)
bdt = BDTsklearn(
model, self.features,
["cls{0}".format(c) for c in range(len(np.unique(self.data_y)))])
bdt.to_tmva("sklearn_ada_multiclass.xml")
bdt.setup_tmva("sklearn_ada_multiclass.xml")
dev = 0.0
for irow in range(self.data_x.shape[0]):
predA = bdt.eval_tmva(self.data_x[irow, :])
predB = bdt.eval(self.data_x[irow, :])
local_dev = np.abs((predA - predB) / predA)
if np.any(local_dev >= 0.05):
print("TMVA: {0}".format(predA))
print("Generalized Tree: {0}".format(predB))
print("Difference: {0}".format(local_dev))
self.assertTrue(np.all(local_dev < 0.05))
dev += local_dev
self.assertTrue(np.all(dev < 0.01))
def test_classify_binary(self):
print("TestBDTsklearn test_classify_binary")
model = GradientBoostingClassifier(n_estimators=10)
model.fit(self.data_x, self.data_y_binary)
bdt = BDTsklearn(model, self.features, ["cls0", "cls1"])
bdt.to_tmva("sklearn_binary.xml")
bdt.setup_tmva("sklearn_binary.xml")
dev = 0.0
for irow in range(self.data_x.shape[0]):
predA = bdt.eval_tmva(self.data_x[irow, :])
predB = bdt.eval(self.data_x[irow, :])
local_dev = np.abs((predA - predB) / predA)
self.assertTrue(local_dev < 0.05)
dev += local_dev
self.assertTrue(dev < 0.01)
def test_regression(self):
print("TestBDTsklearn test_classify_multiclass")
model = GradientBoostingRegressor(n_estimators=10)
model.fit(self.data_x, self.data_y)
bdt = BDTsklearn(model, self.features, ["target"])
bdt.to_tmva("sklearn_regression.xml")
bdt.setup_tmva("sklearn_regression.xml")
dev = 0.0
for irow in range(self.data_x.shape[0]):
predA = bdt.eval_tmva(self.data_x[irow, :])
predB = bdt.eval(self.data_x[irow, :])
local_dev = np.abs((predA - predB) / predA)
self.assertTrue(np.all(local_dev < 0.05))
dev += local_dev
self.assertTrue(np.all(dev < 0.01))
def test_classify_multiclass(self):
print("TestBDTsklearn test_classify_multiclass")
model = GradientBoostingClassifier(n_estimators=10)
model.fit(self.data_x, self.data_y)
bdt = BDTsklearn(
model, self.features,
["cls{0}".format(c) for c in range(len(np.unique(self.data_y)))])
bdt.to_tmva("sklearn_multiclass.xml")
bdt.setup_tmva("sklearn_multiclass.xml")
dev = 0.0
for irow in range(self.data_x.shape[0]):
predA = bdt.eval_tmva(self.data_x[irow, :])
predB = bdt.eval(self.data_x[irow, :])
local_dev = np.abs((predA - predB) / predA)
self.assertTrue(np.all(local_dev < 0.05))
dev += local_dev
self.assertTrue(np.all(dev < 0.01))