Dataset catalogue
This table is generated from the bundled schemas when the documentation is built.
Pass a name to CityLearnEnv to download and cache the dataset on first use; the size
is the download. Default window is the schema’s simulation window, which
simulation_start_time_step, simulation_end_time_step and episode_time_steps can shorten.
Parquet datasets need pip install "citylearn[parquet]", and datasets with LSTM thermal
dynamics need PyTorch; see the installation guide.
Where to start
First runs and controller comparisons:
citylearn_challenge_2022_phase_all_plus_evs, used by the quickstart.Electrical limits, dynamic topology and sub-minute data: the
citylearn_three_phase_*demonstrations; see the features guide.Renewable energy communities with a local market:
pt_rec_4first, thenpt_rec_15,pt_rec_30andpt_rec_100.Demand response and robustness:
citylearn_challenge_2022_phase_all_demand_responseandcitylearn_challenge_2022_phase_all_robustness.
All public datasets
Dataset |
Origin |
Buildings |
EVs |
Step |
Default window |
Format |
Size |
Features |
|---|---|---|---|---|---|---|---|---|
|
Neighbourhood |
4 |
0 |
1 h |
122 days |
CSV |
2 MB |
thermal storage, LSTM thermal dynamics (PyTorch) |
|
Neighbourhood |
100 |
0 |
1 h |
365 days |
CSV |
80 MB |
batteries, thermal storage, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2020 |
9 |
0 |
1 h |
365 days |
CSV |
5 MB |
batteries, thermal storage |
|
CityLearn Challenge 2020 |
9 |
0 |
1 h |
365 days |
CSV |
5 MB |
batteries, thermal storage |
|
CityLearn Challenge 2020 |
9 |
0 |
1 h |
365 days |
CSV |
5 MB |
batteries, thermal storage |
|
CityLearn Challenge 2020 |
9 |
0 |
1 h |
365 days |
CSV |
5 MB |
batteries, thermal storage |
|
CityLearn Challenge 2021 |
9 |
0 |
1 h |
1460 days |
CSV |
19 MB |
batteries, thermal storage |
|
CityLearn Challenge 2022 |
5 |
0 |
1 h |
365 days |
CSV |
3 MB |
batteries |
|
CityLearn Challenge 2022 |
5 |
0 |
1 h |
365 days |
CSV |
3 MB |
batteries |
|
CityLearn Challenge 2022 |
7 |
0 |
1 h |
365 days |
CSV |
3 MB |
batteries |
|
CityLearn Challenge 2022 |
17 |
0 |
1 h |
365 days |
CSV |
7 MB |
batteries |
|
CityLearn Challenge 2022 buildings with demand-response requests |
17 |
0 |
1 h |
365 days |
CSV |
7 MB |
batteries, demand response |
|
CityLearn Challenge 2022 buildings with EVs; the quickstart dataset |
17 |
8 |
1 h |
365 days |
CSV |
13 MB |
EVs, flexible appliances, batteries |
|
CityLearn Challenge 2022 buildings with robustness events |
17 |
0 |
1 h |
365 days |
CSV |
7 MB |
batteries, robustness events |
|
CityLearn Challenge 2023 |
3 |
0 |
1 h |
30 days |
CSV |
< 1 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
3 |
0 |
1 h |
30 days |
CSV |
< 1 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
3 |
0 |
1 h |
92 days |
CSV |
1 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
3 |
0 |
1 h |
92 days |
CSV |
1 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
3 |
0 |
1 h |
92 days |
CSV |
1 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
6 |
0 |
1 h |
92 days |
CSV |
2 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
6 |
0 |
1 h |
92 days |
CSV |
2 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
CityLearn Challenge 2023 |
6 |
0 |
1 h |
92 days |
CSV |
2 MB |
batteries, thermal storage, power outages, LSTM thermal dynamics (PyTorch) |
|
Feature demonstration |
18 |
9 |
15 s |
7 days |
Parquet (pyarrow) |
< 1 MB |
EVs, batteries, phase limits, community market, dynamic topology |
|
Feature demonstration |
18 |
9 |
1 h |
365 days |
CSV |
13 MB |
EVs, batteries, phase limits, community market, dynamic topology |
|
Feature demonstration |
18 |
9 |
15 s |
365 days |
Parquet (pyarrow) |
13 MB |
EVs, batteries, phase limits, community market, dynamic topology |
|
Feature demonstration |
18 |
9 |
1 h |
365 days |
CSV |
13 MB |
EVs, flexible appliances, batteries, phase limits, community market, dynamic topology |
|
Feature demonstration |
17 |
8 |
1 h |
365 days |
CSV |
13 MB |
EVs, flexible appliances, batteries, phase limits, community market |
|
Feature demonstration |
17 |
8 |
15 s |
365 days |
Parquet (pyarrow) |
12 MB |
EVs, flexible appliances, batteries, phase limits, community market |
|
Portuguese renewable energy community (Portugal REC datasets) |
100 |
85 |
15 min |
731 days |
Parquet (pyarrow) |
113 MB |
EVs, flexible appliances, batteries, phase limits, community market |
|
Portuguese renewable energy community (Portugal REC datasets) |
15 |
6 |
15 min |
731 days |
Parquet (pyarrow) |
13 MB |
EVs, flexible appliances, batteries, phase limits, community market |
|
Portuguese renewable energy community (Portugal REC datasets) |
30 |
19 |
15 min |
731 days |
Parquet (pyarrow) |
84 MB |
EVs, flexible appliances, batteries, phase limits, community market |
|
Portuguese renewable energy community (Portugal REC datasets) |
4 |
2 |
15 min |
731 days |
Parquet (pyarrow) |
7 MB |
EVs, flexible appliances, batteries, phase limits, community market |
|
Neighbourhood |
20 |
0 |
1 h |
270 days |
CSV |
330 MB |
LSTM thermal dynamics (PyTorch) |
|
Neighbourhood |
20 |
0 |
1 h |
270 days |
CSV |
330 MB |
LSTM thermal dynamics (PyTorch) |
|
Neighbourhood |
100 |
0 |
1 h |
365 days |
CSV |
80 MB |
batteries, thermal storage, LSTM thermal dynamics (PyTorch) |
|
Neighbourhood |
47 |
0 |
1 h |
365 days |
CSV |
38 MB |
batteries, thermal storage, LSTM thermal dynamics (PyTorch) |
See the dataset reference to build your own dataset.