A time series parser for feeding Numpy arrays to a
Avoids embedding data in the graph as constants.
__init__( data, read_num_records_hint=4096 )
Numpy array input for a
data: A dictionary mapping feature names to Numpy arrays, with two possible shapes (requires keys
VALUESare both vectors of shape [series length] Multivariate;
TIMESis a vector of shape [series length],
VALUEShas shape [series length x number of features]. In any case,
VALUESand any exogenous features must have their shapes prefixed by the shape of the value corresponding to the
read_num_records_hint: The maximum number of samples to read at one time, for efficiency.
Raise an error if the dataset is too small.
Returns a large chunk of the Numpy arrays for later re-chunking.
Tensor versions of the full Numpy arrays.