tf.contrib.timeseries.NumpyReader

Class NumpyReader

Defined in tensorflow/contrib/timeseries/python/timeseries/input_pipeline.py.

A time series parser for feeding Numpy arrays to a TimeSeriesInputFn.

Avoids embedding data in the graph as constants.

__init__

__init__(
    data,
    read_num_records_hint=4096
)

Numpy array input for a TimeSeriesInputFn.

Args:

• data: A dictionary mapping feature names to Numpy arrays, with two possible shapes (requires keys TrainEvalFeatures.TIMES and TrainEvalFeatures.VALUES): Univariate; TIMES and VALUES are both vectors of shape [series length] Multivariate; TIMES is a vector of shape [series length], VALUES has shape [series length x number of features]. In any case, VALUES and any exogenous features must have their shapes prefixed by the shape of the value corresponding to the TIMES key.
• read_num_records_hint: The maximum number of samples to read at one time, for efficiency.

Methods

tf.contrib.timeseries.NumpyReader.check_dataset_size

check_dataset_size(minimum_dataset_size)

Raise an error if the dataset is too small.

tf.contrib.timeseries.NumpyReader.read

read()

Returns a large chunk of the Numpy arrays for later re-chunking.

tf.contrib.timeseries.NumpyReader.read_full

read_full()

Returns Tensor versions of the full Numpy arrays.