def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(data=dataset.train.data),
test=dict(data=dataset.test.data),
target=dict(name=dataset.target.name, classes=dataset.target.values),
columns=[
(
f.name,
(
'object' if f.is_categorical(
strict=False
) # keep as object everything that is not numerical
else 'int' if f.data_type == 'integer' else 'float'))
for f in dataset.features
],
problem_type=dataset.type.
name # AutoGluon problem_type is using same names as amlb.data.DatasetType
)
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
X_train, X_test = impute_array(dataset.train.X_enc, dataset.test.X_enc)
y_train, y_test = dataset.train.y_enc, dataset.test.y_enc
data = dict(train=dict(X=X_train, y=y_train),
test=dict(X=X_test, y=y_test))
def process_results(results):
if results.probabilities is not None and not results.probabilities.shape: # numpy load always return an array
prob_format = results.probabilities.item()
if prob_format == "predictions":
target_values_enc = dataset.target.label_encoder.transform(
dataset.target.values)
results.probabilities = Encoder(
'one-hot', target=False,
encoded_type=float).fit(target_values_enc).transform(
results.predictions)
else:
raise ValueError(
f"Unknown probabilities format: {prob_format}")
return results
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config,
process_results=process_results)
def run(dataset: Dataset, config: TaskConfig):
X_train_enc, X_test_enc = impute(dataset.train.X_enc, dataset.test.X_enc)
data = dict(train=dict(X_enc=X_train_enc, y_enc=dataset.train.y_enc),
test=dict(X_enc=X_test_enc, y_enc=dataset.test.y_enc))
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train_path=dataset.train.path,
test_path=dataset.test.path,
target=dataset.target.name
)
return run_in_venv(__file__, "exec.py",
input_data=data, dataset=dataset, config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(path=dataset.train.path),
test=dict(path=dataset.test.path),
target=dict(index=dataset.target.index),
domains=dict(cardinalities=[0 if f.values is None else len(f.values) for f in dataset.features])
)
config.ext.monitoring = rconfig().monitoring
return run_in_venv(__file__, "exec.py",
input_data=data, dataset=dataset, config=config)
def run(dataset: Dataset, config: TaskConfig):
from amlb.datautils import impute
from frameworks.shared.caller import run_in_venv
X_train_enc, X_test_enc = impute(dataset.train.X_enc, dataset.test.X_enc)
data = dict(train=dict(X_enc=X_train_enc, y_enc=dataset.train.y_enc),
test=dict(X_enc=X_test_enc, y_enc=dataset.test.y_enc))
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(train=dict(path=dataset.train.path),
test=dict(path=dataset.test.path),
target=dict(index=dataset.target.index))
config.ext.monitoring = rconfig().monitoring
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(path=dataset.train.data_path('parquet')),
test=dict(path=dataset.test.data_path('parquet')),
target=dict(
name=dataset.target.name,
classes=dataset.target.values
),
problem_type=dataset.type.name # AutoGluon problem_type is using same names as amlb.data.DatasetType
)
return run_in_venv(__file__, "exec.py",
input_data=data, dataset=dataset, config=config)
def run(dataset, config):
from frameworks.shared.caller import run_in_venv
data = dict(train=dict(X=dataset.train.X, y=dataset.train.y),
test=dict(X=dataset.test.X, y=dataset.test.y),
problem_type=dataset.type.name)
options = dict(serialization=dict(
sparse_dataframe_deserialized_format='dense'))
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config,
options=options)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(X_enc=dataset.train.X_enc, y_enc=dataset.train.y_enc),
test=dict(X_enc=dataset.test.X_enc, y_enc=dataset.test.y_enc),
predictors_type=[
'Categorical' if p.is_categorical(strict=False) else 'Numerical'
for p in dataset.predictors
])
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(
X=dataset.train.X,
y=dataset.train.y
),
test=dict(
X=dataset.test.X,
y=dataset.test.y
)
)
return run_in_venv(__file__, "exec.py",
input_data=data, dataset=dataset, config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
X_train, X_test = dataset.train.X_enc, dataset.test.X_enc
y_train, y_test = unsparsify(dataset.train.y_enc, dataset.test.y_enc)
data = dict(train=dict(X=X_train, y=y_train),
test=dict(X=X_test, y=y_test),
predictors_type=[
'Numerical' if p.is_numerical() else 'Categorical'
for p in dataset.predictors
])
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(data=dataset.train.data, y_enc=dataset.train.y_enc),
test=dict(data=dataset.test.data, y_enc=dataset.test.y_enc),
target=dict(name=dataset.target.name, classes=dataset.target.values),
columns=[(f.name, ('object' if f.is_categorical(
strict=False) else 'int' if f.data_type == 'integer' else 'float'))
for f in dataset.features],
problem_type=dataset.type.name)
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from amlb.datautils import impute_array
from frameworks.shared.caller import run_in_venv
encode = config.framework_params.get('_encode', True)
X_train, X_test = impute_array(
dataset.train.X_enc,
dataset.test.X_enc) if encode else (dataset.train.X, dataset.test.X)
y_train, y_test = (dataset.train.y_enc,
dataset.test.y_enc) if encode else (dataset.train.y,
dataset.test.y)
data = dict(train=dict(X=X_train, y=y_train),
test=dict(X=X_test, y=y_test))
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
data = dict(
target=dataset.target.name,
train=dict(path=dataset.train.path,
X=dataset.train.X,
y=dataset.train.y),
test=dict(path=dataset.test.path, X=dataset.test.X, y=dataset.test.y),
)
options = dict(serialization=dict(
sparse_dataframe_deserialized_format='dense'))
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config,
options=options)
def run(dataset: Dataset, config: TaskConfig):
from amlb.datautils import impute
from frameworks.shared.caller import run_in_venv
data = dict(
train=dict(X=dataset.train.X, y=dataset.train.y),
test=dict(X=dataset.test.X, y=dataset.test.y),
columns=[
(
f.name,
(
'object' if not f.is_numerical(
) # keep as object everything that is not numerical
else 'float')) for f in dataset.predictors
],
problem_type=dataset.type.name)
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
def run(dataset: Dataset, config: TaskConfig):
from frameworks.shared.caller import run_in_venv
train_path = dataset.train.path
test_path = dataset.test.path
backend = config.framework_params.get('_backend')
# ML-Plan requires the target attribute to be the last column
if dataset.target.index != len(dataset.predictors):
train_path = reorder_dataset(dataset.train.path,
target_src=dataset.target.index)
test_path = reorder_dataset(dataset.test.path,
target_src=dataset.target.index)
data = dict(
train=dict(path=train_path),
test=dict(path=test_path),
target=dict(index=dataset.target.index),
)
return run_in_venv(__file__,
"exec.py",
input_data=data,
dataset=dataset,
config=config)
