Is your feature request related to a problem? Please describe.
I want a one-stop solution to easily make use of the EULP dataset to generate a CityLearn dataset.
Describe the solution you'd like
A class in CityLearn for schema generation using EULP dataset.
Makes use of the DOE_XStock library to generate thermal dynamics model training data, building time series and weather data given EULP building IDs.
Function that trains LSTM model with option of using either user-defined hyperparameters, generalized hyperparameters, or fine-tuned hyperparameters. The output should be a .pth file of the trained model's weights and hidden states.
Building schema json with general or user defined values for active actions and observations, DER specification. The rest of the schema is built using the DOE_XStock simulation results and the LSTM results.
Is your feature request related to a problem? Please describe. I want a one-stop solution to easily make use of the EULP dataset to generate a CityLearn dataset.
Describe the solution you'd like
.pth
file of the trained model's weights and hidden states.Describe alternatives you've considered NIL
Additional context NIL