Function reference • modeltime.gluonts

Installation

Use this function to quickly install and configure a GluonTS environment for use with modeltime.gluonts.

install_gluonts() uninstall_gluonts(): Install / Uninstall GluonTS

GluonTS Algorithm Integrations

The main algorithms that have been integrated with modeltime.

deep_ar(): General Interface for DeepAR Time Series Models

deep_state(): General Interface for Deep State Time Series Models

gp_forecaster(): General Interface for GP Forecaster Time Series Models

nbeats(): General Interface for N-BEATS Time Series Models

Saving and Loading Models

The parsnip-adjacent algorithms that implement time series models.

save_gluonts_model() load_gluonts_model(): Saving and Loading GluonTS Models

Dials Tuning Parameters

Tuning parameters for various deep learning models

lookback_length() num_batches_per_epoch() learn_rate_decay_factor() scale_values(): General Tuning Parameters for GluonTS Trainers (All Models)

num_stacks(): Tuning Parameters for NBEATS Models

GluonTS Developer Utilities

These are not meant for general use. Usage is internal to library functionality.

to_gluon_list_dataset(): Convert a data frame to a GluonTS ListDataset

as_pandas_timestamp(): Convert R Date or POSIXt to Pandas Timestamp

is_gluonts_activated() activate_gluonts() get_python_env() check_gluonts_dependencies() check_pytorch_dependencies() detect_default_gluonts_env(): GluonTS Environment Connection Utilities