This documentation is built from the development branch, master. It includes the changes listed under Unreleased, which the latest release, CityLearn 3.6.0, does not have. To read the documentation of the version you installed (pip show citylearn), open the docs folder at its tag on GitHub, for example v3.6.0.

Source code for citylearn.end_use_load_profiles.lstm_model.model_generation_wrapper

import random
from typing import Any, Mapping, Tuple
import numpy as np
import pandas as pd
import torch
from citylearn.end_use_load_profiles.lstm_model.model import LSTM
from citylearn.end_use_load_profiles.lstm_model.model_generation import run

[docs] def run_one_model(config: Mapping[str, Any], df: pd.DataFrame, seed: int) -> Mapping[str, Any]: model, observation_metadata, error_metrics = get_model(config, df, seed) return { 'model': model, 'attributes': { 'hidden_size': config['hidden_size'], 'num_layers': config['num_layer'], 'lookback': config['lb'], 'input_observation_names': observation_metadata['input_observation_names'], 'input_normalization_minimum': observation_metadata['input_normalization_minimum'], 'input_normalization_maximum': observation_metadata['input_normalization_maximum'] }, 'error_metrics': error_metrics }
[docs] def get_model(config: Mapping[str, Any], df: pd.DataFrame, seed) -> Tuple[LSTM, Mapping[str, Any], Mapping[str, float]]: set_random_seeds(seed) lstm, observation_metadata, error = run(config, df) return lstm, observation_metadata, error
[docs] def set_random_seeds(seed: int, benchmark: bool = None, deterministic: bool = None): torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.cuda.manual_seed_all(seed) random.seed(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed_all(seed) torch.backends.cudnn.benchmark = False if benchmark is None else benchmark torch.backends.cudnn.deterministic = True if deterministic is None else deterministic