Curve types
Most of the data in the API is time series data. This page describes the data model and the formats the API returns.
Curves
Section titled “Curves”Time series data is organised into curves. A curve holds the metadata that uniquely identifies the data it contains, for example the Norwegian spot price or a wind production forecast for Denmark. Each curve has a unique ID, a name, and a set of attributes.
There are four curve types. The curve type decides which endpoints you use to read its data.
| Curve type | curve_type value |
Holds | Endpoints |
|---|---|---|---|
| Time series | TIME_SERIES |
One time series | Time series |
| Tagged time series | TAGGED |
One time series per tag | Tagged time series |
| Instance | INSTANCES |
One time series per issue date | Instances |
| Tagged instance | TAGGED_INSTANCES |
One time series per issue date and tag | Tagged instances |
Time series curves
Section titled “Time series curves”The simplest curves. They map directly to a single time series. They are mostly used for actual data, for example historical data where there is a single truth.
Tagged time series curves
Section titled “Tagged time series curves”Like time series curves, but they store multiple time series. Each time series is identified by a tag. A tag can be any string, and you can list the valid tags for each curve. The most common use is ensemble data.
Instance curves
Section titled “Instance curves”Our most used curves. They store multiple time series as instances. An instance is identified by an issue date and usually holds the values of the forecast issued at that date. The API lets you search through instances to find the data you need.
Tagged instance curves
Section titled “Tagged instance curves”The combination of instance curves and tagged time series curves. Each instance is identified by both an issue date and a tag. They are typically used for ensemble forecasts.
Data formats
Section titled “Data formats”The API returns JSON (RFC 7159), with partial support for MessagePack.
{ "id": 809, "name": "pro de-50hz wnd intraday ec00da mwh/h cet min15 f", "frequency": "MIN15", "time_zone": "CET", "curve_type": "INSTANCES", "curve_state": "PUBLIC", "created": "2018-04-10T12:02:36.883+02:00", "modified": "2019-10-02T10:19:56.787+02:00", "issue_frequency": "ANY", "area": "DE-50Hz", "categories": ["PRO", "WND", "INTRADAY"], "commodity": "POW", "unit": "MWh/h", "scenarios": ["EC00DA"], "hasAccess": true, "accessRange": { "begin": "2017-10-31T23:00:00Z", "end": "2019-12-31T23:00:00Z" }, "data_type": "F"}Instance
Section titled “Instance”{ "frequency": "MIN15", "time_zone": "CET", "id": 809, "name": "pro de-50hz wnd intraday ec00da mwh/h cet min15 f", "issue_date": "2018-05-24T09:00:00+02:00", "created": "2018-05-24T09:05:52.148+02:00", "modified": "2018-05-24T09:05:52.148+02:00"}Instance with data
Section titled “Instance with data”{ "frequency": "MIN15", "time_zone": "CET", "id": 809, "name": "pro de-50hz wnd intraday ec00da mwh/h cet min15 f", "issue_date": "2018-05-24T09:00:00+02:00", "created": "2018-05-24T09:05:52.148+02:00", "modified": "2018-05-24T09:05:52.148+02:00", "points": [ [1483477200000, 66493.6135150707], [1483480800000, 65663.09801064023] ]}Time series
Section titled “Time series”{ "frequency": "MIN15", "time_zone": "CET", "id": 1239, "name": "pro dk1 wnd mwh/h cet min15 n", "modified": "2019-09-23T14:03:12.615+02:00", "points": [ [1483477200000, 66493.6135150707], [1483480800000, 65663.09801064023] ]}Time stamps in points are UNIX time stamps in milliseconds: the number of milliseconds since
midnight 1 January 1970 UTC, ignoring leap seconds. In the example above, the two time stamps
are 2017-01-03T22:00:00.000+01:00 and 2017-01-03T23:00:00.000+01:00.
Time series have a strict frequency. An hourly time series has values exactly on each hour, never between.
Access range
Section titled “Access range”You can only read data within your subscription. This limits both which curves you can read
and how far back in history (the from and to dates).
To find your access range for a curve, get the curve by ID or name:
GET https://data-forecasts.volue.com/api/v1/timeseries/curves/get?id=809The response includes accessRange:
{ "accessRange": { "begin": "2017-10-31T23:00:00Z", "end": null }}begin and end are the limits of your access. Query within them with from and to:
GET https://data-forecasts.volue.com/api/v1/timeseries/series/{id}?from=2018-05-18T00:00:00&to=2018-06-07T00:00:00The access range is usually the same for all your curves, unless they come from different
subscriptions. See also /curves/{id}/access.