def test_transform():
valid_values = ["none", "standardize", "normalize", "demean", "descale"]
df = h2o.create_frame(rows=100,
cols=4,
categorical_fraction=0.4,
integer_fraction=0,
binary_fraction=0,
real_range=100,
integer_range=100,
missing_fraction=0,
seed=1234)
model = H2OAggregatorEstimator(target_num_exemplars=5)
try:
for val in valid_values:
model.transform = val
model.train(training_frame=df)
except:
assert False, "Aggregator model should be able to process all valid transform values"
# Try with invalid value
try:
model = H2OAggregatorEstimator(target_num_exemplars=5,
transform="some_invalid_value")
assert False, "Passing invalid value of transform should throw an error"
except:
pass
def test_binary():
df = h2o.create_frame(
rows=1000,
cols=10,
categorical_fraction=0.6,
integer_fraction=0,
binary_fraction=0,
real_range=100,
integer_range=100,
missing_fraction=0.1,
factors=5,
seed=1234
)
params = {
"target_num_exemplars": 100,
"rel_tol_num_exemplars": 0.5,
"categorical_encoding": "binary",
"transform": "normalize"
}
agg = H2OAggregatorEstimator(**params)
agg.train(training_frame=df)
assert agg.aggregated_frame is not None, "Trained model should produce not empty aggregated frame"
assert is_consistent(df.nrows, agg.aggregated_frame), \
"Exemplar counts should sum up to number of training rows"
assert correct_num_exemplars(agg.aggregated_frame, **params), \
"Generated number of exemplars should match target value"
def test_4995():
data = h2o.import_file(DATASET)
raised = False
agg = H2OAggregatorEstimator(model_id="aggregator")
try:
agg.train(training_frame=data)
except H2OResponseError:
raised = True
# H2OAggregatorEstimator.train raises error, when requesting wrong API version
assert raised is False, "Local and Server versions of AggregatorEstimator should match!"
def test_aggregator_effective_parameters():
frame = h2o.create_frame(rows=10000,
cols=10,
categorical_fraction=0.6,
integer_fraction=0,
binary_fraction=0,
real_range=100,
integer_range=100,
missing_fraction=0,
factors=100,
seed=1234)
agg1 = H2OAggregatorEstimator(target_num_exemplars=1000,
rel_tol_num_exemplars=0.5,
categorical_encoding="eigen")
agg1.train(training_frame=frame)
agg2 = H2OAggregatorEstimator(target_num_exemplars=1000,
rel_tol_num_exemplars=0.5)
agg2.train(training_frame=frame)
assert agg2.parms['categorical_encoding']['input_value'] == "AUTO"
assert agg2.parms['categorical_encoding']['actual_value'] == agg1.parms[
'categorical_encoding']['actual_value']
def test_cat_encoding():
valid_values = [
"auto",
"enum",
"one_hot_internal",
"one_hot_explicit",
"binary",
"eigen",
"label_encoder",
"enum_limited",
# "sort_by_response" TODO: This is invalid parameter, remove it
]
df = h2o.create_frame(rows=100,
cols=4,
categorical_fraction=0.4,
integer_fraction=0,
binary_fraction=0,
real_range=100,
integer_range=100,
missing_fraction=0,
seed=1234)
model = H2OAggregatorEstimator(target_num_exemplars=5)
try:
for val in valid_values:
model.categorical_encoding = val
model.train(training_frame=df)
except:
assert False, "Aggregator model should be able to process all valid categorical_encoding values"
# Try with invalid value
try:
model = H2OAggregatorEstimator(
target_num_exemplars=5, categorical_encoding="some_invalid_value")
assert False, "Passing invalid value of categorical_encoding should throw an error"
except:
pass
def test_aggregator_get_mapping_frame():
winequality_df = h2o.import_file(
pyunit_utils.locate("smalldata/wine/winequality-redwhite.csv"))
params = {
"target_num_exemplars": 650,
"rel_tol_num_exemplars": 0.25,
"save_mapping_frame": True
}
agg = H2OAggregatorEstimator(ignored_columns=["quality", "type"], **params)
agg.train(training_frame=winequality_df)
mapping_frame = agg.mapping_frame
assert mapping_frame.names == ["exemplar_assignment"]
assert mapping_frame.nrows == winequality_df.nrows
def test_num_of_exemplars(target_exemplars, tol):
"""Test whether number of generated exemplars corresponds to expected number +/- tolerance
"""
df = h2o.create_frame(rows=10000,
cols=2,
categorical_fraction=0.1,
integer_fraction=0.3,
real_range=100,
seed=1234)
agg = H2OAggregatorEstimator(target_num_exemplars=target_exemplars,
rel_tol_num_exemplars=tol)
agg.train(training_frame=df)
assert agg.aggregated_frame is not None, "Trained model should produce not empty aggregated frame"
assert (1-tol)*target_exemplars <= agg.aggregated_frame.nrows <= (1+tol)*target_exemplars, \
"Final number of aggregated exemplars should be in equal to target number +/- tolerance"
def test_all_params():
data_path = pyunit_utils.locate(
"smalldata/airlines/allyears2k_headers.zip")
df = h2o.import_file(data_path)
params = {
"model_id": "agg",
"training_frame": df,
"response_column": "IsDepDelayed",
"ignored_columns": ["UniqueCarrier"],
"ignore_const_cols": False,
"target_num_exemplars": 500,
"rel_tol_num_exemplars": 0.3,
"transform": "standardize",
"categorical_encoding": "eigen"
}
try:
model = H2OAggregatorEstimator(**params)
model.train(training_frame=df)
except:
assert False, "Should not throw error on valid parameters"
def test_low_cardinality_enum_limited():
raw_data = [
"1|2|A|A",
"1|2|A|A",
"1|2|A|A",
"1|2|A|A",
"1|2|A|A",
"2|2|A|B",
"2|2|A|A",
"1|4|A|A",
"1|2|B|A",
"1|2|B|A",
"1|2|A|A",
"1|2|A|A",
"4|5|C|A",
"4|5|D|A",
"2|5|D|A",
"3|5|E|A",
"4|5|F|A",
"4|5|G|A",
"4|5|H|A",
"4|5|I|A",
"4|5|J|A",
"4|5|K|A",
"4|5|L|A",
"4|5|M|A",
"4|5|N|A",
"4|5|O|A",
"4|5|P|A"
]
raw_data = [el.split("|") for el in raw_data]
df = H2OFrame(raw_data)
agg = H2OAggregatorEstimator(target_num_exemplars=5, categorical_encoding="enum_limited")
agg.train(training_frame=df)
assert agg.aggregated_frame is not None, "Trained model should produce not empty aggregated frame"
assert is_consistent(df.nrows, agg.aggregated_frame), "Exemplar counts should sum up to number of training rows"
# from AggregatorTest.java
assert agg.aggregated_frame.nrows == 7, "Number of exemplars of this test case should be 7"