Quickstart#

Run these commands from a source checkout after Installation. The first two workflows use synthetic data and require no provider accounts.

Generate a fleet dashboard#

uv run --no-sync energy-chile demo --output artifacts/demo --reports

Open artifacts/demo/results/reports/index.html. The pipeline preserves training/holdout boundaries, comparison coverage, input provenance and the synthetic label. It does not download Chilean production data.

Run the applied CCYA case#

uv run --no-sync energy-ccya demo --output artifacts/ccya-demo

This follows the thesis case-study structure: 50 years, seven illustrative paths, four degradation assumptions and 25/50-year comparisons. The applied case guide explains the output files, original data-access limits and the real-data configuration.

Plan a CDS download offline#

from datetime import date
from energy_data_retrieval.sources.cds.era5 import era5_requests

requests = era5_requests(
    date(2020, 1, 1), date(2020, 2, 29),
    variables=["2m_temperature", "surface_solar_radiation_downwards"],
    area=(29.0, -17.0, 28.0, -16.0),  # north, west, south, east
)
print(len(requests))  # two monthly requests; nothing downloaded

Use the provider guide for actual retrieval. Downloads retain native units and timestamps; preparing analysis inputs is an explicit step, especially for accumulated radiation and climate-model calendars.

Next steps#