def test_convert_simple_df_to_tensor():
df = pd.DataFrame({"X1": [1, 3], "X2": [2, 4]})
actual_tensor = convert_pandas_to_tf_tensor(df)
expected_tensor = tf.constant([[1, 2], [3, 4]], dtype=tf.dtypes.int64)
tf.debugging.assert_equal(actual_tensor, expected_tensor)
actual_tensor = convert_pandas_to_tf_tensor(df[["X1"]])
expected_tensor = tf.constant([[1], [3]], dtype=tf.dtypes.int64)
tf.debugging.assert_equal(actual_tensor, expected_tensor)
def test_convert_simple_df_to_tensor_with_dtype():
df = pd.DataFrame({"X1": [1, 3], "X2": [2, 4]})
actual_tensor = convert_pandas_to_tf_tensor(df, dtype=tf.dtypes.float16)
expected_tensor = tf.constant([[1, 2], [3, 4]], dtype=tf.dtypes.float16)
tf.debugging.assert_equal(actual_tensor, expected_tensor)
def test_convert_image_df_to_tensor():
images = np.zeros([4, 3, 32, 32])
df = pd.DataFrame({"image": TensorArray(images)})
actual_tensor = convert_pandas_to_tf_tensor(df)
expected_tensor = tf.zeros([4, 3, 32, 32], dtype=images.dtype)
tf.debugging.assert_equal(actual_tensor, expected_tensor)
def predict(
self,
data: DataBatchType,
feature_columns: Optional[Union[List[str], List[int]]] = None,
dtype: Optional[tf.dtypes.DType] = None,
) -> DataBatchType:
"""Run inference on data batch.
The data is converted into a TensorFlow Tensor before being inputted to
the model.
Args:
data: A batch of input data. Either a pandas DataFrame or numpy
array.
feature_columns: The names or indices of the columns in the
data to use as features to predict on. If None, then use
all columns in ``data``.
dtype: The TensorFlow dtype to use when creating the TensorFlow tensor.
If set to None, then automatically infer the dtype.
Examples:
.. code-block:: python
import numpy as np
import tensorflow as tf
from ray.train.predictors.tensorflow import TensorflowPredictor
def build_model(self):
return tf.keras.Sequential(
[
tf.keras.layers.InputLayer(input_shape=(2,)),
tf.keras.layers.Dense(1),
]
)
predictor = TensorflowPredictor(model_definition=build_model)
data = np.array([[1, 2], [3, 4]])
predictions = predictor.predict(data)
.. code-block:: python
import pandas as pd
import tensorflow as tf
from ray.train.predictors.tensorflow import TensorflowPredictor
def build_model(self):
return tf.keras.Sequential(
[
tf.keras.layers.InputLayer(input_shape=(1,)),
tf.keras.layers.Dense(1),
]
)
predictor = TensorflowPredictor(model_definition=build_model)
# Pandas dataframe.
data = pd.DataFrame([[1, 2], [3, 4]], columns=["A", "B"])
predictions = predictor.predict(data)
# Only use first column as the feature
predictions = predictor.predict(data, feature_columns=["A"])
Returns:
DataBatchType: Prediction result.
"""
if self.preprocessor:
data = self.preprocessor.transform_batch(data)
if isinstance(data, pd.DataFrame):
if feature_columns:
data = data[feature_columns]
tensor = convert_pandas_to_tf_tensor(data, dtype=dtype)
else:
tensor = tf.convert_to_tensor(data, dtype=dtype)
# TensorFlow model objects cannot be pickled, therefore we use
# a callable that returns the model and initialize it here,
# instead of having an initialized model object as an attribute.
model = self.model_definition()
if self.model_weights is not None:
input_shape = list(tensor.shape)
# The batch axis can contain varying number of elements, so we set
# the shape along the axis to `None`.
input_shape[0] = None
model.build(input_shape=input_shape)
model.set_weights(self.model_weights)
prediction = list(model(tensor).numpy())
return pd.DataFrame({"predictions": prediction}, columns=["predictions"])