Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[Breaking] Change default evaluation metric for classification to logloss / mlogloss (#6183)

Browse Source 
* Change DefaultEvalMetric of classification from error to logloss

* Change default binary metric in plugin/example/custom_obj.cc

* Set old error metric in python tests

* Set old error metric in R tests

* Fix missed eval metrics and typos in R tests

* Fix setting eval_metric twice in R tests

* Add warning for empty eval_metric for classification

* Fix Dask tests

Co-authored-by: Hyunsu Cho <chohyu01@cs.washington.edu>
This commit is contained in:
Christian Lorentzen
2020-10-02 21:06:47 +02:00
committed by GitHub
parent e0e4f15d0e
commit cf4f019ed6
18 changed files with 56 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/python/test_basic.py
View File

@@ -81,7 +81,7 @@ class TestBasic(unittest.TestCase):
dtrain = xgb.DMatrix(dpath + 'agaricus.txt.train')
dtest = xgb.DMatrix(dpath + 'agaricus.txt.test')
param = {'max_depth': 2, 'eta': 1, 'verbosity': 0,
'objective': 'binary:logistic'}
'objective': 'binary:logistic', 'eval_metric': 'error'}
# specify validations set to watch performance
watchlist = [(dtest, 'eval'), (dtrain, 'train')]
num_round = 2

8
tests/python/test_basic_models.py
View File

@@ -117,7 +117,8 @@ class TestModels(unittest.TestCase):
# learning_rates as a list
# init eta with 0 to check whether learning_rates work
param = {'max_depth': 2, 'eta': 0, 'verbosity': 0,
'objective': 'binary:logistic', 'tree_method': tree_method}
'objective': 'binary:logistic', 'eval_metric': 'error',
'tree_method': tree_method}
evals_result = {}
bst = xgb.train(param, dtrain, num_round, watchlist,
callbacks=[xgb.callback.reset_learning_rate([
@@ -131,7 +132,8 @@ class TestModels(unittest.TestCase):
# init learning_rate with 0 to check whether learning_rates work
param = {'max_depth': 2, 'learning_rate': 0, 'verbosity': 0,
'objective': 'binary:logistic', 'tree_method': tree_method}
'objective': 'binary:logistic', 'eval_metric': 'error',
'tree_method': tree_method}
evals_result = {}
bst = xgb.train(param, dtrain, num_round, watchlist,
callbacks=[xgb.callback.reset_learning_rate(
@@ -145,7 +147,7 @@ class TestModels(unittest.TestCase):
# check if learning_rates override default value of eta/learning_rate
param = {
'max_depth': 2, 'verbosity': 0, 'objective': 'binary:logistic',
'tree_method': tree_method
'eval_metric': 'error', 'tree_method': tree_method
}
evals_result = {}
bst = xgb.train(param, dtrain, num_round, watchlist,

12
tests/python/test_dmatrix.py
View File

@@ -115,7 +115,9 @@ class TestDMatrix(unittest.TestCase):
eval_res_0 = {}
booster = xgb.train(
{'num_class': 3, 'objective': 'multi:softprob'}, d,
{'num_class': 3, 'objective': 'multi:softprob',
'eval_metric': 'merror'},
d,
num_boost_round=2, evals=[(d, 'd')], evals_result=eval_res_0)
predt = booster.predict(d)
@@ -130,9 +132,11 @@ class TestDMatrix(unittest.TestCase):
assert sliced_margin.shape[0] == len(ridxs) * 3
eval_res_1 = {}
xgb.train({'num_class': 3, 'objective': 'multi:softprob'}, sliced,
num_boost_round=2, evals=[(sliced, 'd')],
evals_result=eval_res_1)
xgb.train(
{'num_class': 3, 'objective': 'multi:softprob',
'eval_metric': 'merror'},
sliced,
num_boost_round=2, evals=[(sliced, 'd')], evals_result=eval_res_1)
eval_res_0 = eval_res_0['d']['merror']
eval_res_1 = eval_res_1['d']['merror']

2
tests/python/test_early_stopping.py
View File

@@ -58,7 +58,7 @@ class TestEarlyStopping(unittest.TestCase):
y = digits['target']
dm = xgb.DMatrix(X, label=y)
params = {'max_depth': 2, 'eta': 1, 'verbosity': 0,
'objective': 'binary:logistic'}
'objective': 'binary:logistic', 'eval_metric': 'error'}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
early_stopping_rounds=10)

5
tests/python/test_with_dask.py
View File

@@ -274,7 +274,7 @@ def test_dask_classifier():
X, y = generate_array()
y = (y * 10).astype(np.int32)
classifier = xgb.dask.DaskXGBClassifier(
verbosity=1, n_estimators=2)
verbosity=1, n_estimators=2, eval_metric='merror')
classifier.client = client
classifier.fit(X, y, eval_set=[(X, y)])
prediction = classifier.predict(X)
@@ -386,6 +386,7 @@ def run_empty_dmatrix_cls(client, parameters):
y = dd.from_array(np.random.randint(low=0, high=n_classes, size=kRows))
dtrain = xgb.dask.DaskDMatrix(client, X, y)
parameters['objective'] = 'multi:softprob'
parameters['eval_metric'] = 'merror'
parameters['num_class'] = n_classes
out = xgb.dask.train(client, parameters,
@@ -482,7 +483,7 @@ async def run_dask_classifier_asyncio(scheduler_address):
X, y = generate_array()
y = (y * 10).astype(np.int32)
classifier = await xgb.dask.DaskXGBClassifier(
verbosity=1, n_estimators=2)
verbosity=1, n_estimators=2, eval_metric='merror')
classifier.client = client
await classifier.fit(X, y, eval_set=[(X, y)])
prediction = await classifier.predict(X)

2
tests/python/test_with_pandas.py
View File

@@ -174,7 +174,7 @@ class TestPandas(unittest.TestCase):
def test_cv_as_pandas(self):
dm = xgb.DMatrix(dpath + 'agaricus.txt.train')
params = {'max_depth': 2, 'eta': 1, 'verbosity': 0,
'objective': 'binary:logistic'}
'objective': 'binary:logistic', 'eval_metric': 'error'}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10)
assert isinstance(cv, pd.DataFrame)