View source on GitHub |
Feature-wise normalization of the data.
Inherits From: PreprocessingLayer
, Layer
, Module
tf.keras.layers.experimental.preprocessing.Normalization(
axis=-1, dtype=None, mean=None, variance=None, **kwargs
)
This layer will coerce its inputs into a distribution centered around 0 with standard deviation 1. It accomplishes this by precomputing the mean and variance of the data, and calling (input-mean)/sqrt(var) at runtime.
What happens in adapt
: Compute mean and variance of the data and store them
as the layer's weights. adapt
should be called before fit
, evaluate
,
or predict
.
Examples:
Calculate the mean and variance by analyzing the dataset in adapt
.
adapt_data = np.array([[1.], [2.], [3.], [4.], [5.]], dtype=np.float32)
input_data = np.array([[1.], [2.], [3.]], np.float32)
layer = Normalization()
layer.adapt(adapt_data)
layer(input_data)
<tf.Tensor: shape=(3, 1), dtype=float32, numpy=
array([[-1.4142135 ],
[-0.70710677],
[ 0. ]], dtype=float32)>
Pass the mean and variance directly.
input_data = np.array([[1.], [2.], [3.]], np.float32)
layer = Normalization(mean=3., variance=2.)
layer(input_data)
<tf.Tensor: shape=(3, 1), dtype=float32, numpy=
array([[-1.4142135 ],
[-0.70710677],
[ 0. ]], dtype=float32)>
Attributes | |
---|---|
axis
|
Integer or tuple of integers, the axis or axes that should be
"kept". These axes are not be summed over when calculating the
normalization statistics. By default the last axis, the features axis
is kept and any space or time axes are summed. Each element in the
the axes that are kept is normalized independently. If axis is set to
'None', the layer will perform scalar normalization (dividing the input
by a single scalar value). The batch axis, 0, is always summed over
(axis=0 is not allowed).
|
mean
|
The mean value(s) to use during normalization. The passed value(s) will be broadcast to the shape of the kept axes above; if the value(s) cannot be broadcast, an error will be raised when this layer's build() method is called. |
variance
|
The variance value(s) to use during normalization. The passed value(s) will be broadcast to the shape of the kept axes above; if the value(s)cannot be broadcast, an error will be raised when this layer's build() method is called. |
Methods
adapt
adapt(
data, reset_state=True
)
Fits the state of the preprocessing layer to the data being passed.
Arguments | |
---|---|
data
|
The data to train on. It can be passed either as a tf.data Dataset, or as a numpy array. |
reset_state
|
Optional argument specifying whether to clear the state of
the layer at the start of the call to adapt , or whether to start from
the existing state. Subclasses may choose to throw if reset_state is set
to 'False'.
|