Skip to main content

Concepts

High level concepts that are used in the Enertel API responses. The 'language of Enertel forecasts'.

Targets​

Users are provided access to forecasts at the target level, which is a given data series (DALMP, RTLMP, RegUp Price) at all of the objects (price nodes, hubs, interfaces) in a given indepedent system operator (ISO). Related terms: target_description, target_id.

Objects​

Objects refer to the entities or items that are being analyzed or forecasted. This could include power plants, price nodes, hubs, interfaces, or system-wide entities. Related terms: object_name, object_id, object_type, iso.

Data Series​

Data series are things being measured over time and can include real time price at specific intervals, day-ahead price, ancillary prices or non-market data like generation by type. Defined as series_name.

Features​

Features are columns of data that are the intersection of a single object and a single data series. The day-ahead price at a given price node is a feature_id that can be queried using our APIs.

Models​

Models are the algorithms and statistical methods used to generate forecasts. They are trained on historical data and are trained for a specific target and are used to generate forecasts for many objects and a given data series. Related terms: model_id, model_created_at, architecture.

Scenarios​

Scenarios represent a dataframe that is a point-in-time snapshot of the inputs used to generate a forecast. They are specific to a target and include all of the relevant future-looking inputs, as they existed at the time of the scenario created_at. Models are run on scenarios to create many sleeves (or 'batches') of forecasts for many different objects and the given target's data series.

Forecast Batches​

Forecast batches (also intuitively called 'sleeves') are groups of forecasts generated at the same time. They are created for one feature_id, from a single model_id, for a single scenario_id. They always contain the same number of forecasts as all other forecast batches for the same target.

Forecasts​

Forecasts are the predictions generated by our models. They include 13 different percentile levels of the forecasted distribution for the given timestamp.