gluonts.model.forecast_generator module#

class gluonts.model.forecast_generator.DistributionForecastGenerator(distr_output)[source]#: Bases: gluonts.model.forecast_generator.ForecastGenerator

class gluonts.model.forecast_generator.ForecastGenerator[source]#

Bases: object

Classes used to bring the output of a network into a class.

class gluonts.model.forecast_generator.QuantileForecastGenerator(quantiles: List[str])[source]#: Bases: gluonts.model.forecast_generator.ForecastGenerator

class gluonts.model.forecast_generator.SampleForecastGenerator[source]#: Bases: gluonts.model.forecast_generator.ForecastGenerator

gluonts.model.forecast_generator.log_once(msg)[source]#

gluonts.model.forecast_generator.make_distribution_forecast(distr, *args, **kwargs) → gluonts.model.forecast.Forecast[source]#

gluonts.model.forecast_generator.predict_to_numpy(prediction_net, args) → numpy.ndarray[source]#

gluonts.model.forecast_generator.recursively_zip_arrays(x) → Iterator[source]#

Helper function to recursively zip nested collections of arrays.

This defines the fallback implementation, which one can specialized for specific types using by doing @recursively_zip_arrays.register on the type. Implementations for lists, tuples, and NumPy arrays are provided.

For an array a (e.g. a numpy array)

_extract_instances(a) -> [a[0], a[1], …]

For (nested) tuples of arrays (a, (b, c))

_extract_instances((a, (b, c)) -> [(a[0], (b[0], c[0])), (a[1], (b
[1], c[1])), …]