TimeSeriesDataFrame.from_iterable_dataset¶

classmethod TimeSeriesDataFrame.from_iterable_dataset(iterable_dataset: Iterable, num_cpus: int = -1) → TimeSeriesDataFrame[source]¶

Construct a TimeSeriesDataFrame from an Iterable of dictionaries each of which represent a single time series.

This function also offers compatibility with GluonTS ListDataset format.

Parameters:

iterable_dataset (Iterable) –

An iterator over dictionaries, each with a target field specifying the value of the (univariate) time series, and a start field with the starting time as a pandas Period . Example:

iterable_dataset = [
    {"target": [0, 1, 2], "start": pd.Period("01-01-2019", freq='D')},
    {"target": [3, 4, 5], "start": pd.Period("01-01-2019", freq='D')},
    {"target": [6, 7, 8], "start": pd.Period("01-01-2019", freq='D')}
]

num_cpus (int, default = -1) – Number of CPU cores used to process the iterable dataset in parallel. Set to -1 to use all cores.

Returns:

ts_df – A data frame in TimeSeriesDataFrame format.

Return type:

TimeSeriesDataFrame