CityLearn.py

import pickle
import datasets
import numpy as np

_DESCRIPTION = """The dataset consists of tuples of (observations, actions, rewards, dones) sampled by agents
    interacting with the CityLearn 2022 Phase 1 environment (only first 5 buildings)"""

_BASE_URL = "https://huggingface.co/datasets/TobiTob/CityLearn/resolve/main"
_URLS = {
    "f_230": f"{_BASE_URL}/f_230x5x38.pkl",
    "LSTM": f"{_BASE_URL}/L_2189x5x4.pkl",
    "RB1": f"{_BASE_URL}/R1_2189x5x4.pkl",
    "RB2": f"{_BASE_URL}/R2_2189x5x4.pkl",
    "Merged1": f"{_BASE_URL}/L_R1_2189x5x8.pkl",
    "Merged2": f"{_BASE_URL}/L_R1_R2_2189x5x12.pkl",
}


class DecisionTransformerCityLearnDataset(datasets.GeneratorBasedBuilder):
    
    # You will be able to load one configuration in the following list with
    # data = datasets.load_dataset('TobiTob/CityLearn', 'data_name')
    BUILDER_CONFIGS = [
        datasets.BuilderConfig(
            name="f_230",
            description="Data sampled from an expert LSTM policy. Sequence length = 230, Buildings = 5, Episodes = 1 ",
        ),
        datasets.BuilderConfig(
            name="LSTM",
            description="Data sampled from an expert LSTM policy. Sequence length = 2189, Buildings = 5, Episodes = 1 ",
        ),
        datasets.BuilderConfig(
            name="RB1",
            description="Data sampled a rule based policy. Sequence length = 2189, Buildings = 5, Episodes = 1 ",
        ),
        datasets.BuilderConfig(
            name="RB2",
            description="Data sampled a rule based policy. Sequence length = 2189, Buildings = 5, Episodes = 1 ",
        ),
        datasets.BuilderConfig(
            name="Merged1",
            description="LSTM + RBC1. Sequence length = 2189, Buildings = 5, Episodes = 1+1 ",
        ),
        datasets.BuilderConfig(
            name="Merged2",
            description="LSTM + RBC1 + RBC2. Sequence length = 2189, Buildings = 5, Episodes = 1+1+1 ",
        ),
    ]

    def _info(self):

        features = datasets.Features(
            {
                "observations": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
                "actions": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
                "rewards": datasets.Sequence(datasets.Value("float32")),
                "dones": datasets.Sequence(datasets.Value("bool")),
            }
        )

        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
        )

    def _split_generators(self, dl_manager):
        urls = _URLS[self.config.name]
        data_dir = dl_manager.download_and_extract(urls)
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                # These kwargs will be passed to _generate_examples
                gen_kwargs={
                    "filepath": data_dir,
                    "split": "train",
                },
            )
        ]

    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
    def _generate_examples(self, filepath, split):
        with open(filepath, "rb") as f:
            trajectories = pickle.load(f)

            for idx, traj in enumerate(trajectories):
                yield idx, {
                    "observations": traj["observations"],
                    "actions": traj["actions"],
                    "rewards": np.expand_dims(traj["rewards"], axis=1),
                    "dones": np.expand_dims(traj.get("dones", traj.get("terminals")), axis=1),
                }