def fit_before_convert(self, dataset):
"""Fit the encoder."""
# If in tf.data.Dataset, must be encoded already.
if isinstance(dataset, tf.data.Dataset):
return
# Convert the data to np.ndarray.
if isinstance(dataset, pd.DataFrame):
dataset = dataset.values
if isinstance(dataset, pd.Series):
dataset = dataset.values.reshape(-1, 1)
# If encoded.
# TODO: support raw string labels for multi-label.
if len(dataset.flatten()) != len(dataset):
if self.num_classes:
self._check_data_shape(dataset.shape[1:])
return
# Fit encoder.
labels = set(dataset.flatten())
if len(labels) < 2:
raise ValueError(
'Expect the target data for {name} to have '
'at least 2 classes, but got {num_classes}.'.format(
name=self.name, num_classes=self.num_classes))
if len(labels) == 2 and not self.multi_label:
self.label_encoder = encoders.LabelEncoder()
else:
self.label_encoder = encoders.OneHotEncoder()
self.label_encoder.fit(dataset)
def fit_before_convert(self, dataset):
# If in tf.data.Dataset, must be encoded already.
if isinstance(dataset, tf.data.Dataset):
if not self.num_classes:
shape = utils.dataset_shape(dataset)[0]
# Single column with 0s and 1s.
if shape == 1:
self.num_classes = 2
else:
self.num_classes = shape
return
if isinstance(dataset, pd.DataFrame):
dataset = dataset.values
if isinstance(dataset, pd.Series):
dataset = dataset.values.reshape(-1, 1)
# Not label.
if len(dataset.flatten()) != len(dataset):
self.num_classes = dataset.shape[1]
return
labels = set(dataset.flatten())
if self.num_classes is None:
self.num_classes = len(labels)
if self.num_classes == 2:
self.label_encoder = encoders.LabelEncoder()
elif self.num_classes > 2:
self.label_encoder = encoders.OneHotEncoder()
elif self.num_classes < 2:
raise ValueError(
'Expect the target data for {name} to have '
'at least 2 classes, but got {num_classes}.'.format(
name=self.name, num_classes=self.num_classes))
self.label_encoder.fit(dataset)
def test_wrong_num_classes_error():
encoder = encoders.OneHotEncoder(num_classes=3)
with pytest.raises(ValueError) as info:
encoder.fit(np.array(["a", "b", "a"]))
assert "Expect 3 classes in the training targets" in str(info.value)
def test_one_hot_encoder_deserialize_transforms_to_np():
encoder = encoders.OneHotEncoder()
encoder.fit(np.array(["a", "b", "a"]))
encoder = encoders.deserialize(encoders.serialize(encoder))
one_hot = encoder.encode(np.array(["a"]))
assert np.array_equal(one_hot, [[1, 0]]) or np.array_equal(one_hot, [[0, 1]])
def test_one_hot_encoder_decode_to_same_string():
encoder = encoders.OneHotEncoder()
encoder.fit(np.array(["a", "b", "a"]))
assert encoder.decode(encoder.encode(np.array(["a"])))[0] == "a"
