liander-open-data

A lightweight SQLite database and Python API for the Liander open data SBI load profiles: 15-minute electricity profiles per SBI code (Dutch Standard Industrial Classification), together with Liander's dictionary of the recommended profile per SBI code.

• One portable .db file holding all profiles and the SBI dictionary, which any SQLite tool can open.
• A small Python API that returns pandas DataFrames.
• A liander-open-data command line tool.
• The SBI dictionary (1,427 codes) ships with the package and works without any download.

Installation

pip install liander-open-data
# optional extras: read the original .xlsx dictionary / export Parquet
pip install "liander-open-data[excel,parquet]"

Quick start

Build the database once. The first time, this downloads the profiles from Liander (~390 MB zip, cached for later) and takes about 1-2 minutes:

liander-open-data build

If you already have the <SBI code>.csv files, pass their folder instead: liander-open-data build path/to/Profiles.

Then query it from Python:

from liander_open_data import ProfileDatabase

with ProfileDatabase("liander_profiles.db") as db:
    db.lookup("8411")                        # dictionary entry for an SBI code
    df = db.get_profile("8411", columns=["mean", "p90"])
    rec = db.recommended_profile("84111")    # Liander's recommended profile
    wide = db.get_profiles(["01", "8411"], column="mean", start="2023-06-01")

Or from the command line:

liander-open-data lookup 8411
liander-open-data search onderwijs --with-profile
liander-open-data export 8411 -o 8411.csv --columns mean p90

Or with plain SQL:

SELECT code, AVG(mean) AS avg_load
FROM profile_timeseries
WHERE timestamp BETWEEN '2023-07-01' AND '2023-08-01'
GROUP BY code ORDER BY avg_load DESC;

Documentation

Full documentation, covering the data model, CLI reference, API reference and the release guide, is in the docs/ folder. To browse it as a website locally, run uv run mkdocs serve.

Development

uv sync --all-groups
uv run pytest
uv run mkdocs serve      # live documentation preview

Data and license

The code is released under the MIT license. The profiles and SBI dictionary are published by Liander N.V. as open data; check Liander's terms before redistributing them.