Datasets:

THULab
/

ree_data

license: cc-by-4.0
task_categories:
  - time-series-forecasting
language:
  - en
  - es
tags:
  - tsfile
  - timeseries
  - electricity
  - energy
  - spain
  - iberia
  - generation
  - demand
  - time-series
  - renewables
  - power-grid
  - format:tsfile
pretty_name: REE Iberian Peninsula Electricity Generation (2008–2026) (TsFile)
size_categories:
  - 1M<n<10M
configs:
  - config_name: default
    data_files:
      - split: ree_data_1
        path: ree_data_1.tsfile
      - split: ree_data_2
        path: ree_data_2.tsfile
      - split: ree_data_3
        path: ree_data_3.tsfile
      - split: ree_data_4
        path: ree_data_4.tsfile

REE Iberian Peninsula Electricity Generation (2008–2026) (TsFile)

Apache TsFile version of daniel-dona/ree-data.

Overview

Electricity demand and generation by source for the Iberian Peninsula, sourced from REE (Red Eléctrica de España). Covers 2008–present at 5-minute resolution, fully deduplicated.

Tables: ree_data
Rows: 1,931,443 across 1 TsFile table(s).
- ree_data: 1,931,443 rows, 2007-12-31 21:00:00 → 2026-05-14 11:15:00 (UTC)

Schema (TsFile structure)

`ree_data.tsfile`

Time (INT64, milliseconds) — the source ts timestamp as INT64 milliseconds.
TAG: none — the table is a single device / single series.
FIELD (37): dem (int64), eol (int64), nuc (int64), gf (int64), car (int64), cc (int64), hid (int64), aut (int64), inter (int64), icb (int64), sol (int64), solFot (int64), solTer (int64), termRenov (int64), cogenResto (int64), expAnd (int64), expMar (int64), expPor (int64), expFra (int64), expTot (int64), impFra (int64), impPor (int64), impMar (int64), impAnd (int64), impTot (int64), conb (int64), turb (int64), gnhd (int64), bat (int64), consBat (int64), dif (int64), vap (int64), year (int32), month (int32), day (int32), hour (int32), date (string)

Conversion notes

The source is a single train split stored as two Parquet shards; both shards are concatenated into the single TsFile ree_data.tsfile (the split is preserved, not mixed with any other split).
Time is the source ts timestamp (hourly, Spanish peninsular time) as INT64 milliseconds. The redundant calendar columns year/month/day/hour/date are kept as FIELDs.
Single device (the Spanish peninsular system), no TAG column; all 38 source columns kept.

Usage

Install the Apache TsFile Python SDK (pip install tsfile) and read a converted file:

from pathlib import Path
from tsfile import TsFileReader

path = Path("ree_data_1.tsfile")
with TsFileReader(str(path)) as reader:
    schemas = reader.get_all_table_schemas()
    print("tables:", list(schemas))
    table_name = next(iter(schemas))
    table = schemas[table_name]
    columns = [column.get_column_name() for column in table.get_columns()]
    print("columns:", columns)
    field_names = [
        column.get_column_name()
        for column in table.get_columns()
        if column.get_column_name() not in {"Time", "time"}
    ]
    if field_names:
        with reader.query_table(table_name, field_names[:3], batch_size=1024) as result:
            batch = result.read_arrow_batch()
            if batch is not None:
                print(batch.to_pandas().head())

Source & license

Original dataset: daniel-dona/ree-data
Source card title / publisher: REE Iberian Peninsula Electricity Generation (2008–2026)
License: cc-by-4.0 (as declared by the original dataset).