Awesome

Elexon Data Portal

The ElexonDataPortal library is a Python Client for retrieving data from the Elexon/BMRS API. The library significantly reduces the complexity of interfacing with the Elexon/BMRS API through the standardisation of parameter names and orchestration of multiple queries when making requests over a date range. To use the ElexonDataPortal you will have to register for an Elexon API key which can be done here.

Installation

The library can be easily installed from PyPi, this can be done using:

pip install ElexonDataPortal

Getting Started

We'll begin by initialising the API Client. The key parameter to pass here is the api_key, alternatively this can be set by specifying the environment variable BMRS_API_KEY which will then be loaded automatically.

from ElexonDataPortal import api

client = api.Client('your_api_key_here')

Now that the client has been initialised we can make a request!

One of the key abstractions within the ElexonDataPortal library is the handling of multiple requests over a date range specified through the start_date and end_date parameters. Each response will be automatically cleaned and parsed, then concatenated into a single Pandas DataFrame. If a settlement period and date column can be identified in the returned data then a new column will be added with the local datetime for each data-point. N.b. that if passed as a string the start and end datetimes will be assumed to be in the local timezone for the UK

start_date = '2020-01-01'
end_date = '2020-01-01 1:30'

df_B1610 = client.get_B1610(start_date, end_date)

df_B1610.head(3)

If you've previously written your own code for extracting data from the Elexon/BMRS API then you may be wondering where some of the normal parameters you pass have gone. The reduction in the parameters passed are due to 4 core drivers:

• Standardisation of date range parameter names
• Removal of the need to specify ServiceType
• Automatic passing of APIKey after client initialisation
• Shipped with sensible defaults for all remaining parameters

The full list of data streams that are able to be requested can be found here. If you wish to make requests using the raw methods these are available through the ElexonDataportal.dev.raw module.

Further information can be found in the Quick Start guide.

What's Changed in v2

The latest release of the library includes a full rewrite of the code-base. We have endeavoured to make the new API as intuitive as possible but that has required breaking changes from v1, if you wish to continue using the historic library use pip install ElexonDataPortal==1.0.4. N.b v1 will not be maintained going forward, you are advised to change over to v2.0.0+.

The key feature changes are:

• Coverage of more BMRS streams
• Automated default values
• Cleaner client API
• A larger range of request types are compatible with the date range orchestrator

Programmatic Library Generation

One of the core features within the ElexonDataPortal library is that it is self-generating, by which we mean it can rebuild itself (including any new API request methods) from scratch using only the endpoints.csv spreadsheet. As well as generating the Python Client library a BMRS_API.yaml file is created, this provides an OpenAPI specification representation of the Elexon/BMRS API. In turn this allows us to automatically generate documentation, as well as run tests on the API itself to ensure that everything is working as expected - during this process we identified and corrected several small errors in the API documentation provided by Elexon.

To rebuild the library simply run the following in the root directory:

python -m ElexonDataPortal.rebuild

N.b. If you wish to develop the library further or use any of the programmatic library generation functionality then please install the development version of the library using:

pip install ElexonDataPortal[dev]

If you are not installing into a fresh environment it is recommended you install pyyaml and geopandas using conda to avoid any dependency conflicts. In future we are looking to release ElexonDataPortal as a conda package to avoid these issues.

Data Stream Descriptions

The following table describes the data streams that are currently retreivable through the API. The client method to retrieve data from a given stream follows the naming convention get_{stream-name}.