gluonts.model.forecast module#

class gluonts.model.forecast.ExponentialTailApproximation(x_coord: List[float], y_coord: List[numpy.ndarray], tol: float = 1e-08)[source]#

Bases: object

Approximate function on tails based on knots and make a inference on query point. Can be used for either interpolation or extrapolation on tails.

Parameters

  • x_coord – x-coordinates of the data points must be in increasing order.

  • y_coord – y-coordinates of the data points - may be a higher numpy array.

  • tol – tolerance when performing the division and computing the log in the exponential extrapolation.

init_exponential_tail_weights() → Tuple[float, float][source]#

Initialize the weight of exponentially decaying tail functions based on two extreme points on the left and right, respectively.

Returns

beta coefficient for left and right tails.

Return type

Tuple

left(x: float) → numpy.ndarray[source]#

Return the inference made on exponentially decaying tail functions.

For left tail, x = exp(beta * (q - alpha)) For right tail, x = 1 - exp(-beta * (q - alpha))

E.g. for x = self.x_coord[0] or self.x_coord[1], return value is exactly self.y_coord[0] or self.y_coord[1], respectively.

Parameters

x – x-coordinate to evaluate the right tail.

right(x: float) → numpy.ndarray[source]#

Return the inference made on exponentially decaying tail functions.

For left tail, x = exp(beta * (q - alpha)) For right tail, x = 1 - exp(-beta * (q - alpha))

E.g. for x = self.x_coord[-1] or self.x_coord[-2] , return value is exactly self.y_coord[-1] or self.y_coord[-2] respectively. :param x: x-coordinate to evaluate the right tail.

tail_range(default_left_tail=0.1, default_right_tail=0.9)[source]#

Return an effective range of left and right tails.

class gluonts.model.forecast.Forecast[source]#

Bases: object

Abstract class representing predictions.

copy_aggregate(agg_fun: Callable)[source]#

Return a new Forecast object with a time series aggregated over the dimension axis.

Parameters

agg_fun – Aggregation function that defines the aggregation operation (typically mean or sum).

copy_dim(dim: int)[source]#

Return a new Forecast object with only the selected sub-dimension.

Parameters

dim – The returned forecast object will only represent this dimension.

dim() → int[source]#

Return the dimensionality of the forecast object.

property freq#
property index: pandas.core.indexes.period.PeriodIndex#
info: Optional[Dict]#
item_id: Optional[str]#
property mean: numpy.ndarray#
property median: numpy.ndarray#
plot(*, intervals=(0.5, 0.9), ax=None, color=None, name=None, show_label=False)[source]#

Plot median forecast and prediction intervals using matplotlib.

By default the 0.5 and 0.9 prediction intervals are plotted. Other intervals can be choosen by setting intervals.

This plots to the current axes object (via plt.gca()), or to ax if provided. Similarly, the color is using matplotlibs internal color cycle, if no explicit color is set.

One can set name to use it as the label for the median forecast. Intervals are not labeled, unless show_label is set to True.

prediction_length: int#
quantile(q: Union[float, str]) → numpy.ndarray[source]#

Compute a quantile from the predicted distribution.

Parameters

q – Quantile to compute.

Returns

Value of the quantile across the prediction range.

Return type

numpy.ndarray

quantile_ts(q: Union[float, str]) → pandas.core.series.Series[source]#
start_date: pandas._libs.tslibs.period.Period#
class gluonts.model.forecast.Quantile(value: float, name: str)[source]#

Bases: object

classmethod from_float(quantile: float) → gluonts.model.forecast.Quantile[source]#
classmethod from_str(quantile: str) → gluonts.model.forecast.Quantile[source]#
name: str#
classmethod parse(quantile: Union[gluonts.model.forecast.Quantile, float, str]) → gluonts.model.forecast.Quantile[source]#

Produces equivalent float and string representation of a given quantile level.

>>> Quantile.parse(0.1)
Quantile(value=0.1, name='0.1')

>>> Quantile.parse('0.2')
Quantile(value=0.2, name='0.2')

>>> Quantile.parse('0.20')
Quantile(value=0.2, name='0.20')

>>> Quantile.parse('p99')
Quantile(value=0.99, name='0.99')
Parameters

quantile – Quantile, can be a float a str representing a float e.g. ‘0.1’ or a quantile string of the form ‘p10’.

Returns

A tuple containing both a float and a string representation of the input quantile level.

Return type

Quantile

value: float#
class gluonts.model.forecast.QuantileForecast(forecast_arrays: numpy.ndarray, start_date: pandas._libs.tslibs.period.Period, forecast_keys: List[str], item_id: Optional[str] = None, info: Optional[Dict] = None)[source]#

Bases: gluonts.model.forecast.Forecast

A Forecast that contains arrays (i.e. time series) for quantiles and mean.

Parameters

  • forecast_arrays – An array of forecasts

  • start_date (pandas._libs.tslibs.period.Period) – start of the forecast

  • forecast_keys – A list of quantiles of the form ‘0.1’, ‘0.9’, etc., and potentially ‘mean’. Each entry corresponds to one array in forecast_arrays.

  • item_id (Optional[str]) – Identifier of the item being forecasted.

  • info (Optional[Dict]) – Additional information that the forecaster may provide e.g. estimated parameters, number of iterations ran etc.

copy_dim(dim: int) → gluonts.model.forecast.QuantileForecast[source]#

Return a new Forecast object with only the selected sub-dimension.

Parameters

dim – The returned forecast object will only represent this dimension.

dim() → int[source]#

Return the dimensionality of the forecast object.

info: Optional[Dict]#
item_id: Optional[str]#
property mean: numpy.ndarray#

Forecast mean.

prediction_length: int#
quantile(inference_quantile: Union[float, str]) → numpy.ndarray[source]#

Compute a quantile from the predicted distribution.

Parameters

q – Quantile to compute.

Returns

Value of the quantile across the prediction range.

Return type

numpy.ndarray

start_date: pandas._libs.tslibs.period.Period#
class gluonts.model.forecast.SampleForecast(samples: numpy.ndarray, start_date: pandas._libs.tslibs.period.Period, item_id: Optional[str] = None, info: Optional[Dict] = None)[source]#

Bases: gluonts.model.forecast.Forecast

A Forecast object, where the predicted distribution is represented internally as samples.

Parameters

  • samples – Array of size (num_samples, prediction_length) (1D case) or (num_samples, prediction_length, target_dim) (multivariate case)

  • start_date (pandas._libs.tslibs.period.Period) – Start of the forecast.

  • item_id (Optional[str]) – Identifier of the item being forecasted.

  • info (Optional[Dict]) – Additional information that the forecaster may provide e.g. estimated parameters, number of iterations ran etc.

copy_aggregate(agg_fun: Callable) → gluonts.model.forecast.SampleForecast[source]#

Return a new Forecast object with a time series aggregated over the dimension axis.

Parameters

agg_fun – Aggregation function that defines the aggregation operation (typically mean or sum).

copy_dim(dim: int) → gluonts.model.forecast.SampleForecast[source]#

Return a new Forecast object with only the selected sub-dimension.

Parameters

dim – The returned forecast object will only represent this dimension.

dim() → int[source]#

Return the dimensionality of the forecast object.

info: Optional[Dict]#
item_id: Optional[str]#
property mean: numpy.ndarray#

Forecast mean.

property mean_ts: pandas.core.series.Series#

Forecast mean, as a pandas.Series object.

property num_samples#

The number of samples representing the forecast.

property prediction_length#

Time length of the forecast.

quantile(q: Union[float, str]) → numpy.ndarray[source]#

Compute a quantile from the predicted distribution.

Parameters

q – Quantile to compute.

Returns

Value of the quantile across the prediction range.

Return type

numpy.ndarray

start_date: pandas._libs.tslibs.period.Period#
to_quantile_forecast(quantiles: List[str]) → gluonts.model.forecast.QuantileForecast[source]#